Spillover effects of carbon, energy, and stock markets considering economic policy uncertainty

This paper uses a two-regime Markov-switching GARCH model to illustrate that the state-switching of returns in the EU carbon market and its associated energy and stock markets is related to the occurrence of major events. Using the European economic policy uncertainty (EEPU) index as the proxy for the turbulence in the economic policies caused by major events, the Granger causality test shows that EEPU is associated with the three markets. To quantitatively analyze the impact of EEPU on the three markets, the GJRGARCH(1,1)-MIDAS model is constructed to solve the mixed frequency problem. The results indicate that EEPU has a negative effect on the long-term volatility of carbon and energy markets, but a positive effect on the stock market. The forecasting performance of the mixed frequency model considering EEPU is better than that of traditional GARCH models. Based on the decomposition of volatility in the mixed frequency model, the spillover index model is applied to analyze the spillover effects of the "carbon-energy-finance" system at three levels: total volatility, short-term volatility, and long-term volatility. The results indicate that at the level of long-term volatility affected by EEPU, the spillover effect of the system fluctuates more frequently and with greater intensity. The energy and stock markets play different roles at different levels of volatility, but the carbon market is always a risk receiver. The pairwise net spillover effects of the energy and stock markets on the carbon market are time-varying and related to the occurrence of major events.