基于银行类指数及其成分股的分析和预测  

The Analysis and Forecast of Bank Index and Its Constituent Stocks

作　　者：韩笑 滕兴虎 窦婷 

机构地区：[1]陆军工程大学,江苏 南京

出　　处：《统计学与应用》2020年第4期506-514,共9页Statistical and Application

摘　　要：以银行类指数及成分股的K线收盘价为研究样本,对银行类指数及其成分股进行了分析。基于银行类指数和其成分股的关系,首先进行了正回归分析,并通过回归诊断筛查出银行类指数与各个成分股样本数据中的异常点。为了使建立的模型更具有实际意义,利用岭估计消除变量间的多重共线性,并基于绝对约束估计和弹性约束估计选择了变量的最优子集,最终给出了符合实际的银行类指数和其成分股之间的线性回归方程。Taking the k-line closing prices of bank indexes and constituent stocks as research samples, the bank index and its constituent stocks are analyzed. Based on the relationship between the bank index and its constituent stocks, a positive regression analysis was carried out, and the abnormal points in the sample data of the bank index and each constituent stock were screened out through regression diagnosis. In order to make the model more practical, ridge estimation is used to elim-inate multicollinearity among variables, and the optimal subsets of variables are selected based on absolute constraint estimation and elastic constraint estimation. Finally, the linear regression equation between the bank index and its constituent stocks is given.

关 键 词：聚类分析 回归分析 岭估计 绝对约束估计 弹性约束估计 

分 类 号：F83[经济管理—金融学]