为了探索期货市场中的非线性特征,选用大连和芝加哥期货市场中黄豆期货价格时间序列,借助配分函数分布图判定期货时间序列存在多重分形特征,并在此基础上利用多重分形谱对期货时序描述多重分形特征.结果表明,与芝加哥期货市场相比,大连期货市场的多重分形强度更大,风险更大.然后对时间序列进行小波变换,将处理后的时间序列的多重分形强度与原始序列进行比较,发现多重分形特征与时间序列中的噪声及长程相关性有关.
In order to test the nonlinearity in futures market, the multifractal characteristics of two future prices were analyzed. Firstly, the distribution diagrams of partition function was applied to diagnose the existence of muhifractal characteristic. Then the characteristics were described by using muhifractal spectra. It points out that compared to the Chicago futures market, the Dalian futures market have stronger muhifractality and greater risk. The original time series and the transformed time series by wavelet transform were compared. It is found that the long-range correlations and the noise in the time series contribute to the muhifractal behavior of the time series.