The time series of cryptocurrency prices provide a unique window into their value and fluctuations.

In this study, an ordinal partition network is constructed using the price signals, and its features are extracted to investigate the variations.

Our research shows that the proposed method indeed works well for analyzing price fluctuations.

We apply the method to ten digital coins, including Bitcoin, Binance coin, and XRP. In particular, the permutation entropy and clustering coefficient are investigated using the minimum, maximum, mean, and the geometric mean functions for the inbound, outbound, and loop directions.

We find that the global clustering coefficient using the minimum function for triplets in a loop in one direction is the best measure in terms of predictive power and insight.

Z. Shahriari, F. Nazarimehr, K. Rajagopal, S. Jafari, M. Perc, M. Svetec, Cryptocurrency price analysis with ordinal partition networks, Applied Mathematics and Computation 430 (2022) 127237.