Computer Engineering and Applications ›› 2020, Vol. 56 ›› Issue (12): 243-249.DOI: 10.3778/j.issn.1002-8331.1903-0233

Previous Articles     Next Articles

High-Dimensional Dynamic Factor Model for Stock Market with Empirical Studies

ZHENG Hongjing, JIANG Mengmeng, ZHOU Jie   

  1. School of Mathematics and Statistics, Xidian University, Xi’an 710126, China
  • Online:2020-06-15 Published:2020-06-09

股票市场的高维动态因子模型及其实证分析

郑红景,蒋梦梦,周杰   

  1. 西安电子科技大学 数学与统计学院,西安 710126

Abstract:

Yield rate and volatility are the most important variables in financial markets. In order to study the rate-influencing factors, the yield rate and volatility model of financial market is established based on the high-dimensional Dynamic Factor Model(DFM). Then this paper introduces the EM algorithm with penalty to estimate sparse parameter of high-dimensional DFM. By applying this model to the stock data of the Shanghai and Shenzhen stock market, the public factors that affect on the yield rate and volatility and the sparse component matrix are obtained. According to the matrix, it is found that there is a common factor in both models which have an effect on most stocks, while others are the industry factors that only impact on a certain industry of the stocks. It is also analyzed why the the factors fluctuate by combining with the domestic relevant policies and events. In addition, the influence of common factor and industry factors are researched to the industry by using the factor contribution rate.

Key words: dynamic factor model, EM algorithm, yield rate, volatility

摘要:

收益率和波动率是金融市场最重要的变量,为研究对其产生影响的因素,建立了收益率和波动率动态因子模型,并引入带惩罚的EM算法得到高维动态因子模型的稀疏参数估计。将此模型应用到沪深交所股票数据中,得到了对股票收益率和波动率产生影响的公共因子及稀疏的因子载荷矩阵。根据因子载荷矩阵,发现在两个模型中都有一个共同因子对绝大多数股票影响,其他因子是对某行业股票产生影响的行业因子。结合国内相关政策和事件等因素,分析了因子波动趋势,并给出了可能的解释。另外,利用因子贡献率,从行业角度分析了共同因子和行业因子对行业股票的影响程度。

关键词: 动态因子模型, EM算法, 股票收益率, 股票波动率