Spillovers of joint volatility-skewness-kurtosis of major cryptocurrencies and their determinants

Spillovers in high-order moments are understudied in the cryptocurrency markets, and notably their joint volatility-skewness-kurtosis spillover effect and its drivers are overlooked. In this paper, we examine dynamics of the spillovers of joint volatility-skewness-kurtosis of major cryptocurrencies and reveal the macroeconomic, financial, and geopolitical factors driving these spillovers. Methodologically, we first use the autoregressive conditional density model to extract the daily higher-order conditional moments of major cryptocurrencies (Bitcoin, Ethereum, Ripple, Litecoin, and Monero) from 9th August 2015 to 23rd September 2021, and then apply the spillover approach to the extracted conditional higher-order moments, namely conditional volatility, conditional skewness, and conditional (excess) kurtosis, in static and dynamic settings. The results show the following: Firstly, cryptocurrencies are interlinked in a time-varying manner by their conditional volatility, conditional skewness, and conditional (excess) kurtosis. Secondly, the spillover index estimated jointly for these three conditional moments exceeds the sum of the three spillover indices estimated for each conditional moment separately, confirming the existence of interdependence across the three conditional moments. Thirdly, inflation expectation, precious metals volatility, and trading volume are significant determinants of the joint volatility-skewness-kurtosis spillover index, irrespective of the pandemic; the investors' focal point switches from the significant corporate spreads before the COVID-19 outbreak to the vital role the term spread plays during the COVID-19 era as a leading indicator of the economic cycle and growth prospect. Financial stress and geopolitical risk are also important, as indicated by the causality-in-quantile analysis. These findings are relevant to investment decisions and policy formulation.