Multiscale spatiotemporal evolution analysis of cryptocurrency returns

Abstract

As the cryptocurrency market rapidly evolves and solidifies its significance within the global financial system, understanding the dynamic characteristics of returns has become increasingly critical. This study examines nine major cryptocurrencies by market capitalization from 1 January 2017, to 22 November 2023, employing the Mann-Kendall breakpoint test, Pettitt test, wavelet analysis, and TVP-VAR model to uncover the trends, breaks, and cross-category correlations of returns. The findings reveal that most cryptocurrencies exhibit no significant return trends. However, the smart contract platform category experienced a pivotal shift in the latter half of 2019, followed by a substantial improvement in market performance. Additionally, smart contract platforms and alternative currencies exhibit similar return patterns, whereas stablecoins remain relatively independent. Nevertheless, during exceptional periods, the trends of all three asset classes tend to converge. Before breakpoints, cryptocurrency correlations generally intensify, while post-breakpoint correlations gradually weaken. During special market events, smart contract platforms exert a pronounced influence, whereas stablecoins maintain stability. This study bridges the gap in long-term return dynamics and cross-category interactions, providing investors with profound insights into market mechanisms and offering policymakers crucial guidance for regulatory and risk management strategies.