Constant vs. Dynamic Hedge Ratios with an Application to Chinese Copper Futures Market

This paper estimates the constant and dynamic hedge ratios from three alternative modeling frameworks, an OLS-based model, a vector error correction model and a multivariate GARCH model, in Chinese copper futures market and investigates their performance using ex post (in-sample) and ex ante (out-ofsample) hedge periods based on the risk-return comparison method. The paper suggests that the dynamic time-varying hedge ratio, compared to alternative constant hedge ratios from OLS regression and VECM model, provide the highest rate of return as well as the greatest portfolio risk reduction over in-sample and out-ofsample hedge periods.