Abstract
The efficiency is the core and foundation of financial markets. Improving the futures market efficiency so as to make the market runs well，is our fundamental purpose of developing futures market. An efficient futures market should have the following characteristics：（1） futures prices and spot prices should remain basically consistent，or with no arbitrage opportunity；（2） the futures market should reflect the available information，and the earnings is non-predictable；（3） there exists no fixed，long-lasting profit opportunities；（4） the futures market does not allow market manipulation exist；（5） the futures market can provide a signal for the macro-economic policy. In this study，a framework to evaluate the futures market efficiency is constructed based this five points，and is used to broadly analyze and evaluate the efficiency of Chinese commodity futures market.
Firstly，whether the futures price of soybean in China is the unbiased prediction to the spot price of soybean has been studied by constructing panel data and using OLS estimator with panel corrected standard errors（PCSE）. The results show that the futures price of soybean in China is the unbiased spot price prediction within 4 months before expiration. At the same time，the history of both the soybean spot price and futures price has an obvious impact to the current spot prices of soybean.
Secondly，empirical study has been carried out to the spot and dynamic relationship between volume and return of Chinese soybean futures market by quantile regression method. The results show that the spot relationship between volume and return of Chinese soybean futures takes a symmetric “V”-shape which yields zero return. This is evidence indicating that good and bad news reacts on almost same degree in soybean futures market. Another result is that the dynamic relationship between volume and return exist in Chinese soybean futures market，which suggests that the lagging-one-day trading volume has impact on current return，presents the same “V”-shape as the spot one. So，the presence of the dynamic relationship between volume and return suggest that soybean futures price has certain predictability.
Thirdly，we tests the presence of the day-of-the-week effect，the quarter effect，the holiday effect on Chinese commodity futures markets returns and conditional variance（volatility）using the GARCH model，taking copper，aluminum，soybean，wheat futures as samples. Results obtained indicate that the returns and volatility of the all futures do not show the day-of-the-week effect as well as and the quarter effect systematicly. However，the volatility of the all futures do show evidence of the holiday effect. Further analysis shows that the inventory adjustment of both long position and short-seller induce the holiday effect of the price volatility. Our evidence is that the open interest decrease significantly on the day before the holiday，and increase significantly on the day after the holiday. The presence of the holiday effect suggest that Chinese futures market and the international futures market has a strong correlation，and Chinese futures market lack the pricing power in the world.
Fourthly，the events of manipulation in Chinese commodity futures market which occurred in recent years are analyzed. The results suggest that less of the futures products，the unreasonable of the investor structure，and inadequate laws and regulations are the main reasons which lead to manipulation，and thus low-efficiency.
Finally，we refer to the Nanhua Futures Index（NFI） as a sample of Chinese commodity futures index，and study the dynamic relationship between NFI and the Consumer Price Index（CPI） of China. We found that NFI can not play a guiding role to Chinese CPI，that is to say that NFI can not provide the reference for the formulation and adjustment of Chinese monetary policy. However，our empirical result shows that American futures index CRB can guide the CPI of USA. The lack of the variety in Chinese futures market and the price-regulation in our country are the main reasons which cause the market not to effectively play its signaling function.
According to the empirical results，we find that there exist many factors which restrict the improving of Chinese futures market efficiency. So，we provide some policy suggestions，such as enforcing information disclosure and education of the investors，breeding the multiple market participants，introducing the efficient stabilization mechanism，increasing the number of the futures products，promoting the marketlization reform further，and so on.
This paper may have several innovation as fallows：（1） it established a realistic and operable framework for evaluation of the information efficiency of commodity futures market；（2） it used OLS estimator with PCSE to study whether the futures price is the unbiased prediction to the spot price；（3） it used quantile regression method to analyze the spot and dynamic relationship between volume and return of futures market；（4） it studied the dynamic relationship between NFI and Chinese CPI for the first time in China.