Exploring cryptocurrency price dynamics and predictability with ordinal networks

Abstract

Ordinal networks represent an innovative and versatile approach for time series analysis, enabling the transformation of data sequences into complex networks based on the relative order of values. This method provides a fresh perspective on uncovering the internal structure of the data, allowing the identification of recurring patterns and predictability dynamics. In our study, we employ ordinal networks and permutation entropy to analyze the predictability and evolving dynamics of four cryptocurrencies: Bitcoin, Ethereum, Litecoin, and Dogecoin. By leveraging this methodology, we investigate the temporal relationships and ordinal transitions that characterize the price fluctuations and volatility of each cryptocurrency, offering deeper insights into their dynamic complexity and predictive potential in cryptocurrency markets.