Ask "Who", Not "What": Bitcoin Volatility Forecasting with Twitter Data

TL;DR

This paper demonstrates that deep learning models using Twitter data, especially author meta-information, significantly improve Bitcoin volatility forecasting accuracy compared to traditional models.

Contribution

It introduces a novel approach combining social media data with deep learning for cryptocurrency volatility prediction and shows the superiority of temporal convolutional networks.

Findings

Temporal convolutional networks outperform classical autoregressive models.

Tweet author meta-information is a stronger predictor than semantic content.

Public dataset of Bitcoin-related tweets is released for future research.

Abstract

Understanding the variations in trading price (volatility), and its response to exogenous information, is a well-researched topic in finance. In this study, we focus on finding stable and accurate volatility predictors for a relatively new asset class of cryptocurrencies, in particular Bitcoin, using deep learning representations of public social media data obtained from Twitter. For our experiments, we extracted semantic information and user statistics from over 30 million Bitcoin-related tweets, in conjunction with 15-minute frequency price data over a horizon of 144 days. Using this data, we built several deep learning architectures that utilized different combinations of the gathered information. For each model, we conducted ablation studies to assess the influence of different components and feature sets over the prediction accuracy. We found statistical evidences for the hypotheses that: (i) temporal convolutional networks perform significantly better than both classical autoregressive models and other deep learning-based architectures in the literature, and (ii) tweet author meta-information, even detached from the tweet itself, is a better predictor of volatility than the semantic content and tweet volume statistics. We demonstrate how different information sets gathered from social media can be utilized in different architectures and how they affect the prediction results. As an additional contribution, we make our dataset public for future research.