Analysis Of The Spillover Effect Among Different Industries In A-shares Market Based On Investor Sentiment

With the reform and opening up and the establishment of a market economy in China, the stock market is improving and more perfect. Till the end of 2013, the total market value of China’s stocks has been accounted for 40.66% of the GDP. Because the stock market plays a key role of the financial system, and it can promote the development of the economic sustainably and healthily, studies on the relevant phenomenon of the stock market have always been a popular academic research in the finance area.Spillover effect, which is a specific form of externalities, is a common phenomenon on the stock market. Spillover effect between industries means that the performance of every industry in the stock market is not only affected to their own factors within the industry, but also affected by factors from other industries’. If the changes caused by spillover occur simultaneous, the performance of the phenomenon is called comovement; if changes occur in different phases, the performance of the phenomenon is called transmission. Spillover effect is a phenomenon of market failure, which can not be fully explained by the classical finance theory based on the assumption of rational people and the efficient market hypothesis(EMH), and it needs to be considered from the perspective of behavioral finance. Investor sentiment, as one of the most important branches of behavioral finance, is been studied and proved by numerous studies that there are interactions between it and the stock price. Therefore, the investor sentiment is considered in this paper, to study what role it plays in the role of spillovers among different industries.The data of stock market is a typical time-series data. Based on the nonlinear, non-stationary characteristics of stock market data, this study used both time domain and frequency domain time series analysis methods. At first, the frequency domain methods is used——to use a combine methods of Ensemble Empirical Mode Decomposition(EEMD) and Symbolic Time Series Analysis(STSA) to extract the volatility characteristics of different time scales of industries’ price index and investor sentiment, and then the time domain methods is used to analyze the different fluctuations.The main empirical results show that: the investor sentiments of industries of A-Shares market constructed in this paper have capabilities of analysis, but different industries have different sensitivities to their investor sentiments; in the short term, there are spillover effects in cyclical industries and chain related industries, while there is no spillover effect among non-cyclical industries. The effects in chain related industries, which is mainly caused by the industrial investor sentiment, is the biggest; in the long-term, there are spillover effects among all industries.