Title: A Dynamic Network Perspective on the Latent Group Structure of Cryptocurrencies
Abstract
The latent group structure in the cryptocurrency market yields information on network risk and dynamics. By forming a dynamic return-based network with coin attributions, we develop a dynamic covariate-assisted spectral clustering method to detect communities. We prove its uniform consistency along the horizons. Applying this new method, we show the return-based network structure and coin attributions, including algorithm and proof types, jointly determine the market segmentation. Based on the network model, we propose a novel “hard-to-value” measure using centrality scores. Further analysis reveals that the group with a lower centrality score exhibits stronger short-term return reversals. Cross-sectional return predictability further confirms the economic meanings of our grouping results and reveal important portfolio management implications.