In this study, we constitute an adaptive hedging method based on empirical mode decomposition (EMD) method to extract the adaptive hedging horizon and build a time series cross-validation method for robust hedging performance estimation. Basing on the variance reduction criterion and the value-at-risk (VaR) criterion, we find that the estimation of in-sample hedging performance is inconsistent with that of the out-sample hedging performance. The EMD hedging method family exhibits superior performance on the VaR criterion compared with the minimum variance hedging method. The matching degree of the spot and futures contracts at the specific time scale is the key determinant of the hedging performance in the corresponding hedging horizon.