结构方程模型研究及其应用

发布时间：2018-02-23 18:15

本文关键词： 结构方程模型涨跌幅度技术指标影响因素模型　出处：《大连海事大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着社会的进步,世界的复杂化,统计科学也在不断向前发展,从最简单的计数方法逐渐发展到可以深层次挖掘数据信息的分析系统,无论哪个阶段都在其相应的历史舞台上扮演着重要角色。如今,在传统的分析方法仍被广泛应用的背景下,人们又提出了新一代的研究方法——结构方程模型。它在传统分析方法的基础上建立了逻辑关系并解决了潜变量的问题,弥补了传统分析在理论假设上的不足并拓宽了传统分析在使用上的限制,为信息处理带来新的模式。因此这项研究被称为"统计学的第三次改革"。技术指标法是预测股票涨跌的一种方式,但预测的同时指标又会对股票直接或间接地产生助涨助跌的影响。本文主要从技术指标法的四大类型指标——人气型指标、超买超卖型指标、大势型指标和趋势型指标出发,分析指标之间存在的因果关系,以此构建技术指标法对股票涨跌幅度影响因素模型,分析各类指标对涨跌幅度的影响效果大小以及它们之间相互影响效果大小。首先,对各类技术指标的特征进行分析构建其对涨跌幅度的影响因素模型,以龙头股份为样本,模型与实际数据拟合后得到的结果显示人气型指标和超买超卖型指标对其涨跌幅度有较大影响。其次,以同样的选股要求将同为服装家纺行业的森马服饰和海澜之家作为样本进行对比研究。验证模型可靠性的同时比较三组模型的拟合结果。最后,运用通达信炒股软件的选股功能,依次将选股条件设定为人气型指标、超买超卖型指标、大势型指标和趋势型指标,分别对它们的使用方法进行编程,运行得到服装家纺行业一年中上涨的股票平均百分占比以及单只股票上涨的平均百分比,进一步验证人气型指标和超买超卖型指标是影响整个服装家纺行业股票涨跌幅度的两个主要因子,并根据论文得到的结论为使用技术指标的投资者们提出了一些建议。
[Abstract]:With the progress of the society and the complication of the world, the statistical science is developing continuously, from the simplest counting method to the analysis system which can mine the data information deeply. No matter which stage plays an important role in its corresponding historical stage. Today, in the context of the traditional analytical methods are still widely used, A new research method, structural equation model, is proposed, which establishes the logic relation and solves the problem of latent variable on the basis of traditional analysis method. It makes up for the deficiency of the traditional analysis in the theoretical hypothesis and widens the limitation of the use of the traditional analysis. New patterns for information processing. So this study is called the "third Reform in Statistics." Technical indicators are a way to predict the rise and fall of stocks. But at the same time, the index can directly or indirectly help the stock to rise and fall. This paper mainly starts from the four types of index of technical index, that is, the index of sentiment, the index of overbuying and overselling, the index of general trend and the index of trend. Based on the analysis of causality among indexes, this paper constructs a model of the influencing factors of the technical index method on the fluctuation range of stock, and analyzes the effect of each index on the range of rise and fall as well as the mutual effect of each other. First of all, the influence of each index on the fluctuation of stock is analyzed. Based on the analysis of the characteristics of various technical indicators, a model of the factors influencing the fluctuation of the technical indicators is established, taking the leading shares as the sample. After fitting the model with the actual data, the results show that the popularity index and the overbought and oversold index have great influence on the rise and fall of the index. Secondly, With the same stock selection requirements, the Sima clothing and Hailan House, which are also in the clothing textile industry, are compared as samples. The reliability of the model is verified and the fitting results of the three groups of models are compared. Finally, Using the stock selection function of the software, the stock selection condition is set as the popular index, the overbought and oversold index, the general trend index and the trend type index in turn, and their usage methods are programmed respectively. Running to get the average percentage of stocks rising in the clothing and home textile industry in one year and the average percentage of individual stocks rising in one year, It is further verified that the index of sentiment and the index of overbuying and overselling are the two main factors that influence the rise and fall of the whole clothing textile industry, and some suggestions are put forward for the investors who use the technical index according to the conclusion of the paper.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：O213

【参考文献】