Risk-adjusted optimization and diversification effects of cryptocurrencies in multi-asset portfolios: a simulation-based and copula-driven approach

Abstract

This study examines the performance characteristics, diversification potential, and volatility of cryptocurrencies relative to traditional financial assets. It employs an integrated approach incorporating mean–variance optimization, Monte Carlo simulation, and copula dependence modeling. The analysis uses daily data for five major cryptocurrencies and several traditional assets from 2015 to 2024. The results show that crypto-only portfolios exhibit weaker risk-adjusted performance than traditional-only portfolios because of their higher risk and negative Sharpe ratio. In mixed portfolios, cryptocurrencies contribute to diversification and risk reduction while remaining a non-dominant component of the overall allocation. Copula modeling revealed symmetric tail dependence during periods of market distress. Moreover, GARCH and EGARCH estimates captured notable volatility clustering and asymmetric shock responses in cryptocurrency markets. In addition, the VAR and Granger causality analyses showed that global uncertainty and market-stress indicators significantly improved the short-run prediction of BTC volatility.