摘要
运用多重分形谱方法对上证指数大幅波动前后各时段高频数据的实证研究发现,不同时段其多重分形谱形状及多重分形谱参数的变化具有一定规律,即指数发生大幅波动时,其多重分形谱图形的开口变至最大,其参数值也有显著性的变化。为了验证上述实证结果的普遍性,提出运用滑动时间窗的思路,将每连续5个交易日的高频数据进行多重分形统计分析,并将其与股指波动进行对比。结果发现,在股价指数发生大幅波动的情况下,多重分形谱参数具有较明显的变化特征,这为进一步描述股票市场的复杂性规律提供了依据。
Based on the multifractal theory, this paper presents an empirical research on the data of high frequency data of Shanghai stock price index before and after fluctuating. It is found that the shape of multifractal spectrum and the parameters of multifractal spectrum follow certain law, namely, when the time series fluctuates sharply, the value of width of multifractal spectrum gets to maximum, and the changes of different time stages are pronounced. In order to verify the significance, by using the concept of sliding window, a sliding window of 240 frequency data in 5 trading days was considered in order to study stock price index fluctuation. The multifractal parameters and coefficients in each window were calculated. It is fond that when the stock price index fluctuates sharply, the parameters are clearly characterized by a strong variability. This has led to a better understanding of complex stock markets.
出处
《系统管理学报》
北大核心
2008年第3期278-282,共5页
Journal of Systems & Management
基金
国家自然科学基金资助项目(70371062)
关键词
大幅波动
多重分形谱参数
滑动时间窗
sharp fluctuation
multifractal spectrum parameters
sliding time window
