Correlation of cryptocurrencies: a dynamic investigation

Abstract

The research line of this paper aims to capture and detect contagion between Bitcoin and the main factors that could have an impact on Bitcoin such as; Ethereum, Ripple, S&P 500, MSCIWorld, MSCIEM50, Gold, VIX, FSI, and new daily cases and deaths due to Covid-19. For such purpose, the paper has been structured in three parts. The first part, aimed to detect the change points in variance from 01/11/2019 to 31/03/2021 using daily data. Main results suggested that Bitcoin change points were: 07/03/2020, 11/03/2020, and 20/03/2020. For Ethereum were: 07/03/2020 and 20/03/2020, and for Ripple were: 7/12/2020, 20/12/2020, and 08/01/2021. These dates coincide with the announcement of COVID-19 virus as a global pandemic (11/03/2020) and the third wave (December 2020 to the 8 of January, 2021). In the second part of the analysis a DCC-MGARCH model was implemented, in which the persistence of volatility, co-movement, and conditional correlation were studied between the different cryptocurrencies and, between Bitcoin, the Equity Indices and Gold. Main results suggested that cryptocurrencies are positively correlated while no correlations were found between Bitcoin and the Equity Indices, nor Bitcoin and Gold. Finally, a Johansen test was done to identify co-integration relationships between Bitcoin, the S&P 500, VIX, FSI and the Covid-19 variables. No co-integration relationships were found. Finally, a Granger causality test was performed among the variables. Main results suggested that the S&P 500 was Granger causing Bitcoin, as well as the VIX. Daily new deaths was Granger causing Bitcoin and new daily cases due to Covid-19. Finally, a relationship was found between new daily cases and new daily deaths.