Early Warning Signals for Cryptocurrency Market States

Abstract

Being archetypal complex systems, financial markets exhibit rich set of dynamics in their interactions. In this paper, we focus on the recently evolved cryptocurrency market as an example of a complex system and analyse the evolution of cross correlation structure of cryptocurrencies in the 5 year period from 2017 to 2022. We observe characteristic correlation structures in the observation time window duration and use these specific structures to cluster the cryptocurrency market in 4 market states.