Study On The Relationship Between The Volatility Of Investor Sentiment And Stock Price Based On The Ensemble Empirical Mode Decomposition

As a major branch of behavioral finance theory, investor sentiment is an important factorto reflect investors’ psychology, and plays a significant role for the operation anddevelopment of the securities market. But relational studies on the impact of investorsentiment and the stock market mainly use traditional econometric models and analyze onlyfrom the time domain. In this context, studying and exploring the relationship betweeninvestor sentiment and stock price fluctuations from the combination of time domain andfrequency domain perspective would perhaps have new discoveries.Based on this background, a new data analysis method called Ensemble Empirical ModeDecomposition (EEMD) is introduced to the study in the issue. From the perspective ofmulti-scale, we use the EEMD to decompose the time series of investor sentiment and stockprices respectively into several separate intrinsic mode functions (IMF) and a residual term.Through this way, we can obtain the fluctuation characteristics of investor sentiment andstock prices sequences at different time scales. And then, we reconstruct the IMFs form highto low frequency in order to obtain the short-term and mid-term fluctuations of the sequences.Remain the residual term as the long trend of the sequences. Finically, combined withtraditional econometric models, we discuss the fluctuation correlation of investor sentimentand stock prices at different time scales.The empirical results show out meaningful conclusions. In the overall effect, investorsentiment and the stock price fluctuations present different fluctuation relations at differenttime scales: in the short term volatility, investor sentiment has a significant positive impact onthe Shanghai Composite price. However, fluctuation of investor sentiment is ahead of theShanghai Composite price fluctuations in mid-term. It‘s reverse in the long term that changesin the price of the Shanghai Composite is ahead of investor sentiment.In the cross-sectional effect, different equity portfolios have different sensitivity to theinvestor sentiment: in the short term, expect the low price-earnings ratio stocks, othercross-section price indexes are significantly affected by market investor sentiment.Meanwhile, the small-cap stocks, the low price stocks, the high price-earnings ratio stocks,the blue chip stocks and low book value stocks are more sensitive to market investorsentiment. In the medium and long term, the performance of the low price stocks, highprice-earnings ratio stocks, blue chip stocks and low book value stocks are more sensitive tomarket investor sentiment. In addition, in the medium and long-term fluctuations of investorsentiment are ahead of the spread sequences, which suggest that investor sentiment has a certain cross-section of explanatory power on the stock market price. All empirical resultsshow that fluctuations of investor sentiment cause the cross-section impact on asset prices,and this difference effects are significant.