معاملات تامینی بازارهای آتی فلز چین

This paper evaluates different hedging strategies for aluminum and copper futures contracts traded at Shanghai Futures Exchange. In addition to usual candidates such as the traditional regression hedge ratio and the hedging strategy constructed from bivariate fractionally integrated generalized autoregressive conditional heteroskedasticity (BFIGARCH) model, two advanced specifications are proposed to account for impacts of the basis on market volatility and co-movements between spot and futures returns. Empirical results suggest that the basis has asymmetric effects and optimal hedging strategy constructed from the asymmetric BFIGARCH model tends to produce the best in-sample and out-of-sample hedging performance.

Copper and aluminum forward/futures contracts began trading in China in 1991. Over the last 15 years, Chinese futures markets have experienced tremendous development. In the early stage, regulators in China struggled with duplicative exchanges and products, large speculative interest, and market manipulations. With implementation of many regulatory reforms on the futures market, the number of exchanges has been reduced to only three and the market has become more efficient, transparent, and integrated into international futures markets in recent years. Currently, copper and aluminum futures contracts can only be traded at Shanghai Futures Exchange (SHFE), one of three remaining exchanges in China, and their prices display a certain degree of integration with the prices of copper and aluminum futures traded in London Metal Exchange (LME) (see, for example, Zhang (2003)) and therefore can be used to predict spot price movements.1 Using daily price data of September and December 1994 copper futures contracts, Tung (1997) showed that the variability of the basis (i.e., difference between spot and futures prices) is smaller than the variability of the spot price, suggesting that copper futures market can provide a hedging tool to domestic users. Although China's State Council has given approval for eight Chinese nonferrous metal producers to engage in hedging in overseas futures markets (Platt's Metals Week, Vol. 75, Issue 33, page 6), the number is limited as the government only grants approvals to companies with large output capacities and good performance. Therefore, a more careful study of optimal hedging strategy and hedging effectiveness using futures contracts traded in the SHFE is needed for the majority of domestic producers and consumers. Academic studies of Chinese copper and aluminum futures markets are limited, and in particular we can identify no studies of hedging strategies and hedging effectiveness for these markets when we initiated the current research. This paper presents the first attempt to fill in the gap. Several hedging strategies including naïve strategy, constant strategy, and different dynamic strategies are investigated for Chinese copper and aluminum futures. In addition to usual candidates such as the traditional regression hedge ratio and the hedging strategy constructed from bivariate fractionally integrated generalized autoregressive conditional heteroskedasticity (BFIGARCH) model, two advanced specifications are proposed to account for impacts of basis on market volatility and co-movements. Empirical results suggest that basis has asymmetric effects on market behaviors. Moreover, optimal hedging strategy constructed from the asymmetric BFIGARCH model tends to produce the best in-sample and out-of-sample hedging performance. The remainder of the paper is organized as follows. The immediately following section discusses the development of copper and aluminum futures trading in China. Thereafter, optimal hedging strategy and several model specifications are presented, including the OLS regression model, the BFIGARCH model, and the symmetric and asymmetric BFIGARCH models. We then turn to data descriptions and provide preliminary data analysis. The following section presents the estimated results for each model and discusses their implications. From which we construct dynamic hedged portfolios and evaluate hedging performance accordingly. The conclusion is given in the final section.

In this paperwe evaluate different hedging strategies for aluminumand copper futures contracts traded at the Shanghai Futures Exchange. To provide the most useful information to market participants, several statistical models are considered to generate optimal hedging strategies, including the traditional regression hedge ratio, the hedging strategy constructed from the bivariate fractionally integrated generalized autoregressive conditional heteroskedasticity (BFIGARCH) model, and two advanced specifications which are intended to account for impacts of the basis on market volatility and co-movements. Empirical results suggest that the basis has asymmetric effects on the market behaviors. Specifically, spot and futures markets behave differently when the basis is positive as compared to when the basis is negative. This finding provides important implications for futures hedging. The optimal hedge strategy based upon the asymmetric BFIGARCH model tends to produce the best hedging performance for both insample and out-of-sample comparisons among several competing models.