燃料油期货市场成交量、持仓量与波动性关系

采用GARCH模型、方差分解以及脉冲响应函数等方法研究了我国燃料油期货市场的价格波动与成交量和持仓量之间的关系.通过研究得到以下结论: 成交量对价格波动具有很强的解释作用,可以根据上一期成交量的变动, 预测下一期的价格波动;当期持仓量对价格波动具有很强的吸收作用, 持仓量增大时,期货价格波动将随之减小,但滞后期持仓量的变动对期货价格的波动性不具有解释作用,因此无法根据持仓量的历史数据预测未来的价格波动.同时考虑成交量和持仓量时, 当期成交量表现出很强的解释作用,若当期成交量增大, 则价格波动也将增大,反之则减小;当期持仓量在成交量和持仓量同时增加时对价格波动的影响小于在成交量增加、而持仓量没有增加时对价格波动的影响;而滞后期的成交量和持仓量对价格波动均没有显著影响.价格波动的残差扰动大部分由其自身造成;波动性对持仓量的影响比较明显;而成交量与持仓量之间存在显著的相互影响. 脉冲响应曲线表明,燃料油期货市场符合混合分布假设, 市场深度的价格效应是暂时的.

Relations among volume, open interest and volatility in fuel oil futures market

As the study on interior structure and operating characters was deficient in China fuel oil futures market, the dynamic relationships among volatility, volume and open interest were empirically studied through GARCH model, variance decompose, impulse response function and so on. The results show that volume play important roles in explaining volatility. One can forecast volatility by the variability of past volume. The current open interest can explain volatility too; the volatility would decrease if the current open interest increases. On the contrary, the lagged open interest performs no function in explaining the volatility, that is to say, it is impossible to forecast the future price by historical open interest. Considering volume and open interest simultaneously, the volatility would increase if current volume increases, conversely, the volatility would decrease if current volume decreases. The function to volatility that current open interest performs when trading volume and open interest increase together is lesser than when volume increases while open interest doesn't increase. Furthermore, the lagged volume and the lagged open interest perform nothing to volatility. In variance decomposition, residual disturbance of volatility is mainly caused by itself; the volatility has effect on open interest; the volume and open interest have obvious effect on each other. In the end, the impulse response graphs imply that our fuel oil futures market is in accord with the mixture of distributions hypothesis, and the price effect of market depth is temporal.