论文部分内容阅读
我国股市个股价格同时上涨或同时下跌的联动现象极为普遍,传统上使用向量自回归、协整、有向非循环图等方法主要用于少量股票或市场之间的联动性研究,不适于直接对大规模个股之间的联动关系进行研究。文章关注大规模时序图模型结构建立及估计方法,通过将ADL方法引入SPACE算法,提出了可以估计高维低样时序图模型的ADL-SPACE算法;设计模拟实验考察了算法中惩罚参数λ值的设置对于节点自回归相关性捕获的有效性;在实证研究中,文章使用了ADL-SPACE算法对个股联动研究了三方面的内容:1.基于个股联动的代表性行业之间的联动性;2.设计了我国A股市场中行业联动强度,对行业内外联动性进行综合评价和分析;3.采用一阶滞后个股基于时序图模型结果构造了投资组合,模拟显示收益预期表现良好。以上研究均表明时序SPACE图模型方法在大规模股票的联动探测中有较好的应用前景。
The phenomenon that the price of individual stocks in China’s stock market goes up or down at the same time is extremely common. The traditional methods such as vector autoregression, cointegration and directed acyclic graph are mainly used for the research of the linkage between a few stocks or the market, Large-scale stocks between the linkage between the research. This paper focuses on the establishment of large-scale timing diagram model structure and estimation methods. By introducing the ADL method into the SPACE algorithm, an ADL-SPACE algorithm that can estimate the high-dimensional and low-order timing diagram model is proposed. The simulation experiment is conducted to investigate the value of the penalty parameter λ In the empirical study, the article uses the ADL-SPACE algorithm to study the linkage of individual stocks in three aspects: 1. The linkage between representative industries based on individual stock linkage; 2 The strength of industry linkage in A-share market of China is designed, and the internal and external linkage of the industry is comprehensively evaluated and analyzed.3. The first-order lagged stocks are used to construct the investment portfolio based on the timing diagram model. The simulation shows that the expected return is good. The above studies show that the sequential SPACE graph model has good application prospects in the linkage detection of large-scale stocks.