The causal relationship between returns and trading volume in cryptocurrency markets : recursive evolving approach

This chapter examines the time-varying causal relationship between trading volume and returns in cryptocurrency markets. The chapter employs a novel Granger causality framework based on a recursive evolving window procedure. The procedures allow detecting changes in causal relationships among time series by considering potential conditional heteroskedasticity and structural shifts through recursive subsampling. The chapter analyzes the return-volume relationship for Bitcoin and seven other altcoins: Dash, Ethereum, Litecoin, Nem, Stellar, Monero, and Ripple. The results suggest rejecting the null hypothesis of no causality, indicating bi-directional causality between trading volume and returns for Bitcoin and the altcoins except Nem and Stellar. The findings also highlight that the causal relations in cryptocurrency markets are subject to change over time. The chapter may conclude that trading volume has predictive power on returns in cryptocurrency markets, implying potential benefits of constructing volume-based trading strategies for investors and considering trading volume information in developing pricing models to determine the fundamental value of the cryptocurrencies.