Dynamic Dependence and Tail Risk in Technology, Cryptocurrency and Commodity Markets

This study examines the evolution of dependence structures and tail risk transmission among technology equities, Bitcoin, Gold, and Crude Oil during 1 January 2016–1 January 2026. The analysis focuses on NVIDIA (NVDA), AMD, Tesla (TSLA), Bitcoin (BTC), Gold and Oil, covering major disruptions including the COVID-19 pandemic and the Russia–Ukraine conflict. An integrated methodological framework combines DCC-GARCH modeling, R-vine copulas, tail dependence analysis, complexity measures and machine learning-based forecasting techniques. The findings reveal volatility persistence and time-varying correlations, especially between technology equities and BTC during crisis periods. Regime analysis reveals that dependence structures are not stable in time. Lower-tail dependence intensifies during periods of market stress, indicating increased downside risk transmission. Gold remains weakly connected to the other assets, while Bitcoin has the strongest exposure to extreme downside co-movements. Complexity analysis based on the Scale-Dependent Lyapunov Exponent (SDLE) indicates heterogeneous dynamics across scales, characterized by local divergence and stabilization at broader scales. Forecast results based on Random Forest and XGBoost models provide limited predictive gains over benchmark specifications, suggesting that dependence and tail risk modeling offer better insight than short-horizon return predictions. These results are important for monitoring tail risk transmission for financial stability policies.