Empirical analysis of cryptocurrency price movements and their relationship to other assets : a VAR & rolling VaR approach

Abstract

Cryptocurrency is a relatively new digital instrument often referred to "digital gold," yet our understanding of its behaviour remains limited. The scarcity of both data and research makes it difficult to study its dynamics and implications which poses challenges for this emerging asset class, especially when compared to gold, which benefits from extensive scholarly attention. Existing studies in circulation delve into the interconnectedness of Bitcoin with various asset classes such as commodities, currencies, and Equities focusing on themes such as volatility spillovers, returns, and hedging strategies. These papers utilize modelling techniques such as DY models, GARCH models, regression analysis, and VAR-based models in which most papers do not consider the efficiency of different cryptocurrencies along with the dynamic relationships in a post- recession period & economic expansion period, resulting in literature gaps. A blend of methodologies was implemented within this research to enhance the analysis effectiveness. This incorporates a mixed methodology approach of VAR, Rolling VaR, the Diebold-Yilmaz Connectedness Framework, Rolling VAR, GFEVD, Granger Causality, IRF’s and Correlation Analysis with respect to volatility spillovers, dynamic correlation patterns, and responses to economic shocks across various periods. The findings confirm Hypothesis 1, showing volatility spillovers between cryptocurrencies, blockchain equities are high with the DY suggesting a higher spillover effect in periods of systematic events such as COVID 19 & lower spillover during post-recession periods. Hypothesis 2 is partially accepted, as dynamic correlation being lower post-recession with Granger causality results showing Blockchain Equities holding predictive power of cryptos for 3-4 Year periods & Other Cryptos holding predictive power for LT (7 Years). Hypothesis 3 is partially accepted, indicating cryptocurrencies shock reaction are most impactful with other sectoral-adjacent assets & recessionary periods containing more instability and volatility. Impacts in post- Recession periods for non-crypto adjacent was identified to be lower. These results hold regulatory importance in terms of highlighting the potential for volatility mitigation strategies in cryptocurrencies. Policymakers may consider promoting legislation in favour of stablecoins, regulating specific cryptocurrency disclosures & regulating Defi Finance as if they were banks. Additionally, the research could serve as a valuable guide for individual and institutional investors in managing their cryptocurrency based on economic cycles & interpreting them as a hedge or a diversifier accordingly.