Comovement and Instability in Cryptocurrency Markets

We analyze the correlations of daily price returns for nine major cryptocurrencies between April 2013 and November 2018 and estimate their evolution using bivariate and multivariate modelling approaches. We detect pronounced time variation and fid these correlations to be generally increasing between early 2017 and late 2018. We then adopt a right-tail variation of the Augmented Dickey-Fuller unit root test to identify and date-stamp periods of mildly explosive behavior (statistical instability) in the time series of the Network Value to Transactions (NVT) ratio (a measure of the dollar value of cryptocurrency transaction activity relative to its network value) of six cryptocurrencies. We show statistically significant evidence of mild explosiveness in all of them. At the end of 2017 and in 2018, several major cryptocurrencies experience significant (often simultaneous) instability associated with rising NVT ratios. Instability is a steady feature of cryptocurrency markets.