基于集成经验模态分解的投资者情绪与股价波动的关系研究

发布时间：2018-03-13 09:10

本文选题：投资者情绪　切入点：股价　出处：《华南理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：投资者情绪作为行为金融理论的一大分支，是反映投资者心理的重要因素，对证券市场的运行和发展起着非常重要的作用。但目前有关情绪对股市影响的研究，多采用传统计量模型，仅从时间维度进行了分析。在此背景下，从时间域和频率域相结合的角度深入分析投资者情绪变化与股票价格波动的影响关系，或许有新的发现。基于此，本文将一种新的数据分析方法——集成经验模态分解（Ensemble EmpiricalMode Decomposition，EEMD）引入到该问题的研究中。从多尺度的角度出发，运用EEMD方法分别将投资者情绪和股价序列分解成若干个独立的、不同尺度的本征模态函数（Intrinsic Mode Function，IMF）分量和一个残余项，分别提取出序列在不同时间尺度下的波动特征。然后将得到的IMFs按照高低频重构为序列的短期波动项，中期重大事件影响项，保留残余项作为序列的长期趋势项，再结合计量模型探讨了在不同时间尺度下投资者情绪对股价的总体影响效应和横截面影响效应。实证结果表明，就总体效应而言，投资者情绪与股指价格波动在不同时间尺度下呈现出不同的波动关系。从短期波动来看，投资者情绪对上证综指价格波动具有显著正向影响；中期投资者情绪波动领先于上证综指价格波动，而长期则转变为上证综指价格领先投资者情绪波动。就横截面效应而言，不同特征股票组合对投资者情绪的敏感度不同。从短期来看，除了低市盈率股票以外，其他横截面股指价格序列均显著受到市场投资者情绪的影响，且小盘股、低价股、高市盈率股票、绩优股及低市净率股票对市场投资者情绪更为敏感；从中长期来看，则表现为低价股、高市盈率股票、绩优股及低市净率股票对市场投资者情绪更为敏感；且在中长期下投资者情绪领先于价差序列的波动，这表明投资者情绪对股票市场横截面价格具有一定的解释能力，说明投资者情绪对资产价格的波动产生了横截面的影响，且这种影响具有显著的差异性。
[Abstract]:As a branch of behavioral finance theory, investor sentiment is an important factor reflecting investor psychology and plays a very important role in the operation and development of securities market. The traditional econometric model is mostly used, only from the time dimension. Under this background, it is possible to analyze the relationship between investor sentiment change and stock price fluctuation from the angle of time domain and frequency domain, and there may be some new findings. Based on this, a new data analysis method-Ensemble EmpiricalMode decomposition (EEMD) is introduced to the study of this problem. The EEMD method is used to decompose investor sentiment and stock price sequences into several independent, different scale intrinsic Mode function (IMF) components and a residual term. After extracting the fluctuation characteristics of the sequence at different time scales, the obtained IMFs is reconstructed into the short-term fluctuation term of the sequence according to the high and low frequency, the influence item of the medium-term major event, and the residual term as the long-term trend item of the sequence. Combined with the econometric model, the overall and cross-sectional effects of investor sentiment on stock price in different time scales are discussed. The empirical results show that, as far as the overall effect is concerned, investor sentiment and stock index price fluctuate in different time scales. Investor sentiment has a significant positive impact on the price volatility of the Shanghai Composite Index. In the case of cross-sectional effects, investor sentiment volatility in the medium term is ahead of the price fluctuation in the Shanghai Composite Index, while in the long run it has turned into a leading investor sentiment fluctuation in the Shanghai Composite Index. The sensitivity of different characteristic stock combinations to investor sentiment is different. In the short term, except for low price-earnings ratio stocks, other cross-section stock index price sequences are significantly affected by market investor sentiment, and small-cap stocks, low-priced stocks, The stocks with high price-earnings ratio, high price-to-earnings ratio and low price-to-book ratio are more sensitive to the sentiment of market investors, while in the medium and long term, they are characterized by low price stocks, high price-earnings ratio stocks, high transcripts stocks and low price-to-book ratio stocks, which are more sensitive to investors' sentiment in the market. In the medium and long term, investor sentiment leads to the fluctuation of spread sequence, which indicates that investor sentiment has a certain ability to explain the cross-section price of stock market, which indicates that investor sentiment has a cross-sectional influence on the fluctuation of asset price. And this kind of influence has remarkable difference.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F830.59;F830.91;F224

【参考文献】