An Empirical Study On The Risk Of EU ETS Carbon Market Based On Dynamic Dependence

As an effective platform for global climate response,the international carbon trading market also provides investors with more alternative financial assets.EUA and CER have become the two most important trading commodities in the carbon futures market.However,due to the impact of global economic situation,political policies and energy weather,carbon futures asset prices fluctuate dramatically.Studying the dynamic interdependence and the measurement of risks will help investors make better investment decisions between the two futures markets.From March 14,2008 to July 28,2017,2410 sets of daily price data of EUA and CER,two of the most representative Carbon Asset futures contracts in the international carbon trading market,were selected from the European Climate Exchange(ECX)as the research objects to analyze the dynamic dependence,covering the first 5 years of whole second and third stages of futures trading.In this paper,using Bai-Perron endogenous structural mutation test method to divide the third stage of carbon market into two periods with March 24,2015 as breakpoint.First,VAR model is constructed for four periods,and linear,nonlinear Granger causality test,impulse response and variance decomposition are combined to explain the forerunner relationship between EUA and CER markets.After determining the existence and orientation of the relationship,the GARCH-Copula model is used to fit the two-asset series of three short-terms,and then the optimal Copula function is selected to study the dynamic dependence structure of the two markets respectively.Based on the Copula function,the EUA futures assets and their investment groups with CER assets are calculated by Monte Carlo simulation method.The combined VaR value is used to analyze the degree of risk dispersion in the current international carbon futures market,which provides a more accurate theoretical basis for the risk control of the carbon market and the portfolio of carbon financial derivatives.The result shows that there is a one-way or two-way leading relationship between the two in the three periods,and the level of linkage decreases gradually;What’s more,the fitting effect of EUA sequence with ARMA-GARCH-t model is better,while the CER sequence is more inclined to ARMA-GJRGARCH-N model to express asymmetric fluctuations;Last but not least,the optimal Copula function in the three periods is t-Copula,Gumbel Copula and time-varying Rotated Gumbel Copula in turn,and the tail correlation coefficient decreased significantly;Finally,under 95%confidence,the risk of EUA and CER portfolio is significantly lower than that of a single EUA asset,and the risk reduction ratio is large,so the investment risk can be reduced to a certain extent by incorporating CER assets into the investment category.These conclusions provide a decision-making basis for the portfolio strategy based on carbon futures.