Energy markets and CO2 emissions: Analysis by stochastic copula autoregressive model

• We examine the dependence between the volatility of CO2 emissions and energy prices.
• The dependence will be measured by the dynamic SCAR copula pair by pair.
• To model the marginal distributions of the variables, we use the GAS model.
• To evaluate high dimensional integral in the likelihood function we use EIS method.
• The dependence is dynamic and varies over time especially in period of crisis.

We examine the dependence between the volatility of the prices of the carbon dioxide “CO2” emissions with the volatility of one of their fundamental components, the energy prices. The dependence between the returns will be approached by a particular class of copula, the SCAR (Stochastic Autoregressive) Copulas, which is a time varying copula that was first introduced by Hafner and Manner (2012) [1] in which the parameter driving the dynamic of the copula follows a stochastic autoregressive process. The standard likelihood method will be used together with EIS (Efficient Importance Sampling) method, to evaluate the integral with a large dimension in the expression of the likelihood function. The main result suggests that the dynamics of the dependence between the volatility of the CO2 emission prices and the volatility of energy returns, coal, natural gas and Brent oil prices, do vary over time, although not much in stable periods but rise noticeably during the period of crisis and turmoils.