摘要
运用相空间重构技术,分别对上海黄金交易所和伦敦黄金交易市场的现货黄金价格时间序列进行相空间重构,运用互信息法计算嵌入延迟,运用Cao方法计算嵌入维数,运用小数据量法计算最大Lya-punov指数,运用G-P算法估算关联维数和Kolmogorov熵,得出上海黄金市场具有混沌和分形特征的结论。进而对两系统的混沌特性、有效价格预测时间尺度等进行比较,发现两市场整体的混沌程度相当,其中上海黄金市场的复杂度小于伦敦黄金市场,但其价格波动的发散性更强、最大可预测时间更短。最后指出,形成此差异的原因可能是上海黄金交易所是一个次要市场,其价格形成机制不同于作为主要市场的伦敦黄金市场。
Using the phase space reconstruction technique(PSRT),this paper reconstructs the phase spaces of gold price time series of Shanghai gold market and London gold market.And it uses the mutual information method to select delay time,and uses the Cao method to calculate the minimum embedding dimension,and uses the method for small data set to calculate the largest Lyapunov exponent,and uses the G-P method to calculate the correlation dimensions and Kolmogorov entropies of these two markets.It draws the conclusion that there exist the chaos character in the gold price time series of SGE gold exchange.Then it compares the chaos characters and the available forecast time scales of these two markets,and considers that the reason for this difference is that SGE is not a main market,and its price formation mechanism is different from that of London gold market.
出处
《技术经济》
CSSCI
2012年第10期126-132,共7页
Journal of Technology Economics
基金
国家社会科学基金项目"发展循环经济与中国特色的再制造产业协调发展研究"(10BGL010)
