Cryptocurrency Day Trading and Framing Prediction in Microblog Discourse

Abstract

With 56 million people actively trading and investing in cryptocurrency online and globally in 2020, there is an increasing need for automatic social media analysis tools to help understand trading discourse and behavior. In this work, we present a dual natural language modeling pipeline which leverages language and social network behaviors for the prediction of cryptocurrency day trading actions and their associated framing patterns. This pipeline first predicts if tweets can be used to guide day trading behavior, specifically if a cryptocurrency investor should buy, sell, or hold their cryptocurrencies in order to make a profit. Next, tweets are input to an unsupervised deep clustering approach to automatically detect trading framing patterns. Our contributions include the modeling pipeline for this novel task, a new Cryptocurrency Tweets Dataset compiled from influential accounts, and a Historical Price Dataset. Our experiments show that our approach achieves an 88.78% accuracy for day trading behavior prediction and reveals framing fluctuations prior to and during the COVID-19 pandemic that could be used to guide investment actions. 1 Introduction Beginning with the 2008 introduction of Bitcoin (BTC) (Nakamoto, 2008), a cryptocurrency for a Peer-to-Peer cash system, the use of cryptocurrencies and their corresponding blockchains have increasingly gained in popularity. In 2019, the number of Americans owning cryptocurrency doubled from 7% in 2018 to 14%, representing about 35 million people trading and investing with cryptocurrency (Partz, 2019). This increase is largely due to the capability of cryptocurrency to improve various applications ranging from increased security of smart contracts to facilitating less expensive, faster cross-border international payments. Another contributing factor to this growth is that digital coins fulfill the prop-042 erty of storing value similar to other fiat currencies, 043 which are government-issued currencies not backed 044 by physical commodities, e.g., the American dollar 045 or euro. Finally, cryptocurrency popularity can be 046 associated with its high day trading volume. As of 047 January 2021, the combined worth of all cryptocur-048 rencies was $1 trillion 1 , with Bitcoin accounting 049 for $650 billion of this amount. To put this in per-050 spective, the average trading volume of Amazon 051 Inc. is $13 billion per day -less than one-fifth of 052 the BTC daily volume of $70 billion. 2 053 Cryptocurrencies were born on the internet, 054 gained their visibility through online and social 055 media coverage, and many investors follow the 056 advice of well-known cryptocurrency experts on 057 Twitter to guide their personal investment strate-058 gies (Mone, 2019). Because cryptocurrency prices 059 can fluctuate quickly, resulting in real-life financial 060 gains or losses, models that can rapidly analyze 061 trending discourse on Twitter can be harnessed to 062 guide and benefit investors. 063 Additionally, work in computational linguistics 064 and the social sciences have shown the benefit of 065 studying framing, which is how someone discusses 066 a topic in order to influence or alter the opinion of 067 the public, for understanding microblog discourse 068