摘要
以国内各期货品种日收盘价格序列作为研究对象,对中国期货市场是否存在混沌进行全面的检验。首先运用新的最大交易量复权法对期货价格数据进行采样,再进行收益率和对数线性去趋势平稳化处理,运用R/S分析和BDS检验来检验其非线性,运用递归图方法进行确定性检验,发现国内期货市场普遍具有非线性和确定性。其后,对这些时间序列进行相空间重构,计算最大Lyapunov指数、关联维数和Kolmogorov熵等几何不变量,从而得出中国期货市场具有混沌和分形特征的结论,为进一步的混沌预测分析打下基础。
We investigate the nonlinearity of Chinese futures market by using the method of R/S,BDS,and investigate its determinism by using the recurrence plots.The prices time series is sampled with a new max-volume restoration method,then is dealt by using two de-trend methods to meet the requirement of weak stationary.It is found that the market is driven by nonlinear and deterministic mechanisms.Then,using the Phase Space Reconstruction Technique(PSRT) method,we reconstructed the phase spaces of the de-trended prices time series,and calculated the largest Lyapunov exponents,correlation dimensions and Kolmogorov entropies.At last,the evidence of the existence of chaos is found in Chinese futures market.
出处
《系统工程》
CSSCI
CSCD
北大核心
2012年第1期43-53,共11页
Systems Engineering
基金
国家社科基金资助项目(10BGL010)
教育部人文社会科学基金资助项目(09YJA630064)