Cryptocurrencies : a copula based approach for asymmetric risk marginal allocations

Given the increasing interest in cryptocurrencies shown by investors and researchers, and the importance of the potential loss scenarios resulting from investment/trading activities, this research provides market operators with a dynamic overview on the short-term portfolio tail risk contribution of six widely-traded cryptocurrencies. Considering the high volatility dynamics of the cryptocurrency market, realized volatility measures computed from different frames (1m, 5m, 15m, 30m, 1h) are included in the estimation of univariate GARCH models, to be used in combination with copula functions for VaR/ES Monte Carlo simulations. Even if results lack data frequency ordinality in terms of out-of-sample goodness, Bitcoin and Litecoin are generally recognized as the safest and riskiest currency respectively on an equally-weighted framework, reflecting how the contribution to portfolio returns is not representative of the real grade of risk diversification.