The Nonlinear Connectedness Among Cryptocurrencies Using Markov-Switching VAR Model

Abstract

This paper aims to examine the regime-dependent dynamic relations among the leading cryptocurrencies using the Markov Switching Vector Autoregressive model. Our findings suggest evidence in favor of regime-switching properties in the cryptocurrency market. Furthermore, these findings provide strong evidence in favor of nonlinear connectedness among cryptocurrencies; and, thus, it is necessary to employ the MS-VAR model to determine the dynamic nonlinear connectedness between cryptocurrencies. Moreover, we find that the degree of connectedness and volatility spillover is different in both regimes and, based on the transition probability matrix, the low volatility regime is more lengthy than the high volatility regime.