Exogenous drivers of Bitcoin and Cryptocurrency volatility – A mixed data sampling approach to forecasting

• Published in: Journal of international financial markets, institutions & money institutions & money, 2019-11, Vol.63, p.101133, Article 101133
• Main Authors: Walther, Thomas, Klein, Tony, Bouri, Elie
• Format: Article
• Language: English
• Subjects: Bitcoin; Cryptocurrencies; GARCH; Mixed data sampling; Volatility
• ISSN: 1042-4431; 1873-0612
• DOI: 10.1016/j.intfin.2019.101133

•We use the GARCH-MIDAS framework to identify drivers of Cryptocurrency volatility.•In contrast to former studies, we use an out-of-sample setting only.•We find the Global Real Economic Activity to be the best explanatory variable for long-term Cryptocurrency volatility.•Especially during bear markets, explanatory variables offer better predictions than the simple GARCH.•Averaging explanatory variables shows promising results. We apply the GARCH-MIDAS framework to forecast the daily, weekly, and monthly volatility of five highly capitalized Cryptocurrencies (Bitcoin, Etherium, Litecoin, Ripple, and Stellar) as well as the Cryptocurrency index CRIX. Based on the prediction quality, we determine the most important exogenous drivers of volatility in Cryptocurrency markets. We find that the Global Real Economic Activity outperforms all other economic and financial drivers under investigation. We also show that the Global Real Economic Activity provides superior volatility predictions for both, bull and bear markets. In addition, the average forecast combination results in low loss functions. This indicates that the information content of exogenous factors is time-varying and the model averaging approach diversifies the impact of single drivers.