Multiscale dependence and risk contagion between European carbon market, energy, and financial markets

This paper aims to investigate dependency structures and risk spillovers at various time scales between European carbon, energy, stock, and exchange markets. The return series of the markets are decomposed into multi-scale sequences by the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) approach. The ARFIMA-GARCH-skew t model is proposed to estimate the marginal distributions of the return series at raw, short-, and long-term scales. Then, C-Vine copula and CoVaR are integrated to examine their dependency structures and risk spillovers. The findings indicate that the risk spillovers from EUA to the electricity market are the greatest, for both upside and downside risk. Additionally, the upside and downside risk spillovers between EUA and other markets tend to be asymmetric at the raw and short-term scales, while the differences at the long-term scale decrease. Others being equal, the risk spillover effects are stronger at the raw and short-term scales compared to the long-term scale. This work provides valuable insights for investors with different investment horizons regarding systematic risk contagion mechanisms among the European carbon market, energy, and financial markets. It contributes them to grasp risk sources, transmission paths and influence degrees of carbon markets accurately, and provide an important basis for formulating effective short- and long-term risk management strategies. • The dependency structures between markets are heterogeneous under different time scales. • The risk spillovers from EUA to the electricity market are the greatest, no matter the upside or downside risk. • The risk spillovers at the raw and short-term scales are stronger than those at the long-term scale.