The paper selected the carbon emission trading yields data from 2014 to 2017 in Shenzhen. A generalized autoregressive conditional heteroscedasticity model was used to find the best way to hedge the risk of spot carbon emissions in Shenzhen carbon emission trading exchange market. The variances of carbon spot and coal futures were first examined. The dynamic hedging rate was calculated too. The results showed that according to the actual data and market change strategies, the dynamic hedging rate is better than the optimal hedging rate that can hedge risk better. The carbon emission trading yield was found to exhibit aggregate fluctuation; in addition, the dynamic hedging can better hedge risk timely than static hedging. This provides investors with the basis for decision-making to hedge risk in carbon emission trading and helps investors to maximize their returns under certain risk conditions.