Quantifying the randomness of the stock markets
Quantifying the randomness of the stock markets
复制标题
DOI：
10.1038/s41598-019-49320-9
复制
发表时间：
2019-09-04
影响因子：
4.6
通讯作者：
Delgado-Bonal, Alfonso
中科院分区：
文献类型：
--
作者：
Delgado-Bonal, Alfonso
随机性已被数学定义和量化的时间序列使用算法，如近似熵（近似熵）。尽管近似熵独立于任何模型，可以用于任何时间序列，但由于市场具有不同的统计值，它不能直接应用于金融数据序列之间的比较。在本文中，我们进一步开发使用近似熵来量化不断变化的数据序列中的模式的存在，定义一个措施，允许使用最大熵方法的时间序列和时代之间的比较。我们将该方法应用于股票市场，作为其应用的一个例子，显示出根据经济形势，与适应性市场假说一致，六个分析市场的模式数量发生了变化。
Randomness has been mathematically defined and quantified in time series using algorithms such as Approximate Entropy (ApEn). Even though ApEn is independent of any model and can be used with any time series, as the markets have different statistical values, it cannot be applied directly to make comparisons between series of financial data. In this paper, we develop further the use of Approximate Entropy to quantify the existence of patterns in evolving data series, defining a measure to allow comparisons between time series and epochs using a maximum entropy approach. We apply the methodology to the stock markets as an example of its application, showing that the number of patterns changed for the six analyzed markets depending on the economic situation, in agreement with the Adaptive Markets Hypothesis.