High frequency multiscale relationships among major cryptocurrencies: portfolio management implications
Walid Mensi, Mobeen Ur Rehman, Muhammad Shafiullah, Khamis Hamed Al‐Yahyaee, Ahmet Şensoy
Abstract
This paper examines the high frequency multiscale relationships and nonlinear multiscale causality between Bitcoin, Ethereum, Monero, Dash, Ripple, and Litecoin. We apply nonlinear Granger causality and rolling window wavelet correlation (RWCC) to 15 min-data. Empirical RWCC results indicate mostly positive co-movements and long-term memory between the cryptocurrencies, especially between Bitcoin, Ethereum, and Monero. The nonlinear Granger causality tests reveal dual causation between most of the cryptocurrency pairs. We advance evidence to improve portfolio risk assessment, and hedging strategies.
Topics & Concepts
CryptocurrencyGranger causalityPortfolioEconometricsEconomicsCausality (physics)Nonlinear systemWaveletComputer scienceFinancial economicsPhysicsArtificial intelligenceQuantum mechanicsComputer securityBlockchain Technology Applications and SecurityMarket Dynamics and VolatilityComplex Systems and Time Series Analysis