The Relationship Between Order Imbalance And Individual Future Returns And Volatility Of Agricultural Futures Market In China

Compared with the international futures market, agricultural futures market in China though started late, but now has become one of the most important parts of derivatives market and financial markets in China. Therefore, how to find the trend from market data, in other words,the classification for the direction of the order to study order imbalance, defined with buy-initiated trades minus the sellers initiated trades, which not only is the important index in the stock market, in the futures market, is also equally important.In this paper, we study the relationship between the order imbalance with the returns and volatility of the agricultural futures market of China.Before that, we need to measure the order imbalance, that is to say,we should choose a trade classification algorithm which is suitable for agricultural futures market in China.This article selects two popular algorithms, one is the traditional algorithm Lee- ready(LR) algorithms,and the other one is the bulk volume classification algorithm(BVC), presented in 2012, and compare the accuracy of them with application of the internal real futures data of system provided by Dalian Commodity Exchange.The results found that the accuracy of LR algorithm is much higher, almost all over 80%, some even as high as 90% above, however, the accuracy of BVC algorithm is hovering around 70%.Obviously, LR algorithm is applied well and even better in agricultural futures market in China,compared with the results from the foreign market researches.Therefore,we select the LR algorithm to measure the order imbalance. Futher more,we find that the order imbalance exists significantly positive autocorrelated.Then, we estimate the relationship between order imbalances and individual futures returns through the analysis of the time series regression firstly, with the data of three kinds of agricultural futures products in 2011, soybean meal,soybean oil and palm oil from Dalian Commodity Exchange.We find that there are extremely significant positive correlations between order imbalances and futures returns,and under controlling the contemporaneous order imbalances, the lagged order imbalances is negatively related with the futures returns.Beyond that,we also find that order imbalances does not exist the prediction on the futures returns.In the following exploration for order imbalances impact on volatility, we find that in the future market,order mbalances has a significant negative impact on contemporaneous volatility.That means, when buy-initiated trades are much more than sell-initiated trades in the futures market, namely the order imbalance is positive, then the order imbalance is more worse and volatility is more slight.On the contrary, when the sellinitiated trades dominated in the market, the volatility increases much more with order imbalances are much worse.Similarly, the results show that lagged order imbalance has no power of prediction for agricultural futures volatility.