Cryptocurrencies have become a prominent investment tool recently with increasing interest in them and their relationships with stock and foreign exchange markets. We analyze here the cross-correlations of price changes of different cryptocurrencies using Random Matrix Theory and extract community structures by constructing minimum spanning trees, finding their eigenvalues contrast sharply with universal predictions of Random Matrix Theory. We reveal distinct transient community structures among different groupings of cryptocurrencies. By studying the cross-correlation dynamics of sub-communities we find evidence of collective behaviour. Furthermore, we compare eigenvalue changes and find prominent groupings following a community trend, useful for creating cryptocurrency portfolios.

随着对加密货币及其与股票和外汇市场的关系越来越感兴趣，加密货币已成为一种重要的投资工具。我们在这里使用随机矩阵理论分析不同加密货币的价格变化的相互关系，并通过构建最小生成树来提取社区结构，发现其特征值与随机矩阵理论的普遍预测形成鲜明对比。我们揭示了加密货币的不同分组之间不同的瞬态社区结构。通过研究子社区的互相关动态，我们发现了集体行为的证据。此外，我们比较特征值变化并发现社区趋势后的突出分组，这对创建加密货币投资组合很有用。