AI-Carbon-Energy: Spillover effects and drivers in interconnected markets

This study explores the spillover effects between the AI market, international carbon market, and energy markets based on a time-varying parameter vector autoregression model. Further, the study uses a multivariate quantile-to-quantile regression model to identify the macro and micro factors influencing spillover effects. The results show that there are significant spillover effects among the three markets. The AI and new energy markets are the main risk-transmitting markets with respect to the return and volatility spillovers. For skewness and kurtosis, all traditional energy markets except the gas market become risk transmitters. Across all moments, the carbon market consistently is a net recipient of risk. The spillover effects clearly vary with time. Short-term dynamics drive returns and skewness connections, and long-term effects primarily drive volatility and kurtosis connections. In addition, geopolitical risk, economic policy uncertainty, climate risk, AI technological progress, and investor attention may exert differentiated impacts on various spillover effects.