运用重标极差分析方法(R/S分析法)、消除趋势波动分析(DFA方法)以及多重分形消除趋势波动分析方法(MFDFA方法),对我国螺纹钢线材市场收益率序列进行实证研究。结果表明:我国螺纹钢和高线两种钢材的日收益序列具有尖峰厚尾特征,并不服从正态分布,存在显著的正相关性和长记忆性;且螺纹钢和高线的日收益率序列具有明显的多重分形特征,因此仅用单一的标度指数对其进行描述是不合适的。
Using fractal rescaled range analysis method (the R/S method), detrended fluctuation analysis (the DFA method) and multi-fractal detrended fluctuation analysis (MF-DFA), we conduct an empirical study on the time series of returns on the rebar and wire rod prices in China. We find that the time series of returns on the rebar and wire rod prices in China are characterized with leptokurtic and heavy tailed. The time series are not normally distributed and show significant positive correlation and long-term memories, indicating that the time series of the rebar and wire rod markets have obvious multi-fractal properties. The scale variations of the time series show that a single-scale index is insufficient to describe the commodity price fluctuation. These results may lead to a better understanding of complex Chinese steel spot markets.