干散货航运价格指数分形结构研究

A Fractal Structure Study on Shipping Price Index of Dry Bulk

运用分形分布理论对国际干散货航运价格指数的分布结构进行研究,对价格指数序列的Hurst指数及盒维数进行估算,结果表明干散货航运价格指数时间序列具有尖峰厚尾及长期记忆性等分形分布特征。通过对分形分布与正态分布拟合结果的比较发现,正态分布会高估时间序列的收益,并低估其风险,而分形分布能够更好地对其进行拟合,本文对运价指数序列的分形分布参数进行了估算。

Based on the theory of fractal distribution, this paper studies the distribution structure of international shipping price index of dry bulk, and estimates the Hurst index and the box di- mension of the price index series. The result shows that the time series of the shipping price index of dry bulk is characterized by such fractal features as fat-tail distribution and long-term memory. Meanwhile, the comparisons between the fitting results of the fractal distribution and the normal distribution show that normal distribution may over-estimate the income of time series and underestimate its risks; while fractal distribution can fit it better, as well as estimating the fractal structure index of shipping price index series.