Deciphering crypto interconnections using a multi-analytical approach

Abstract

Blockchain technology has primarily underpinned cryptocurrencies that are used either as speculative investment vehicles or for transaction facilitation. There has been a keen interest in understanding the dynamics of interconnectedness and conditional correlations among cryptocurrency prices. While most studies have primarily focused on Bitcoin or the top few cryptocurrencies, this study adopts a comprehensive, multi-analytical approach, incorporating other smaller cryptos that appeal to small and medium investors. Pearson correlational analysis explores the interconnectedness among cryptos and investigates co-movement in crypto prices through their returns, volatility, volume traded, and the CCi30 index returns. Principal Component Analysis (PCA) is used to identify highly correlated clusters, summarizing cross-sectional information based on covariance within the predictors. The predictive regression model of Granger Causality test is applied as a vector autoregression (VAR) forecasting method to examine Granger causality of price movements within the clusters identified. The findings from the correlational matrices of returns and volatilities show no difference in behaviours between larger and smaller cap ones, whereas correlations in trading volumes indicate high correlations in large market-caps. Smaller market-cap cryptos exhibit stronger correlations in volatilities than the larger market cap ones. Two highly correlated clusters emerged from the PCA analysis, with Binance Coin (BNB) and Ripple (XRP) exhibiting greater influence than Bitcoin (BTC) and Tether (USDT) in the second cluster. The findings will enable cryptocurrency users and investors to grasp price mechanisms better, offering valuable insights to improve their decision-making abilities.