Analyzing clustered factors in the cryptocurrency market with Random Matrix Theory

Abstract

The cryptocurrency market is a dynamic and complex system. Factor models can identify latent factors that systematically influence asset returns and are useful for unraveling such complexities. The latent factors can represent the underlying economic, financial, or investor behavioral phenomena driving the price movements of cryptocurrencies. In this paper, we approach the problem from the perspective of Random Matrix Theory (RMT) and assume that while some factors affect all cryptocurrencies, some others are cluster-specific. In particular, we distinguish between stablecoins and non-stablecoins. We find that there are up to two global factors for cryptocurrencies. The results at the cluster level highlight that stablecoins are affected by a larger number of factors than standard cryptocurrencies.