Theoretical And Empirical Research On The Behavior Of Half-informed Traders In The Stock Market

Investors are the main body of market. Investors’ trade behavior is the basic power of market operation. So the analysis of the different classification of investors and their trading activity is especially important in the field of financial research. When different investors making their own trading strategy market information is the fundamental basis of trading. The micro-structure of the market transaction data as the information carrier not only contains the rich market information, but also includes investors trade behavior, even including the investor sentiment and trading strategy more in-depth information. The market investors simply divided into two categories, informed traders and uninformed traders have apparently not corresponding to the increasingly rich investors transaction data contains information. Therefore, through the research of financial data classification and its trading behavior to develop traders’ in-depth digging, enriching the financial results and deeper understanding of the various phenomenons in the securities market has positive significance. The specific content mainly embodied in the following respects.The first part is the concept of half-informed traders group and proves the existence. In the study of financial engineering, information for investors and market research has been carried out under the hypothesis of dual, namely the market exists only informed and the two types of traders. In this paper, we build the game model and we will assume a more generalized and found that only the two types of existing market traders can’t reach the state of the continuous equilibrium can only arrive at instantaneous equilibrium at some point in time. And the true market can be in a state of long-term equilibrium, based on the findings presented in this paper and a half-informed trader group. It exists between informed traders and uninformed traders as a fuzzy group. So we combined with the dynamical system and the differential equation stability theories prove the existence of the group in the market.The main idea of the second part is to build the probability of half-informed trading model and empirical study. Based on the theoretical basis of informed trader and their own characteristics, the investment model of informed traders do not have absolute advantage on the market information. Therefore stand in the perspective of information can be referred to as a "community". The classic EKOP model is also under the hypothesis of dual build and estimate the probability of informed trading(PIN), so this article combined with the method of fuzzy mathematics theory, the EKOP model extended to fuzzy space and fuzzy EKOP model is established, the model can effectively estimate that half the size of the probability of informed trading(HPIN).After model establishment, in order to study the characteristics of different market environment and investors on HPIN size, we make a comparative empirical study of the influence of the selection of the mainland market and the Hong Kong market. The data on the selected in dual-listed AH as the research sample, effectively avoid the otherness of different industries and listed companies, to reduce the error in the process of research. Through the comparative study found that the size of the HPIN under different market environment has significant differences.In the third part, we focus on the research of HPIN relations with different market phenomenon. From the macro market phenomenon, there is mainly embodied in the "flock effect" and the market crowd behavior such as public opinion. From microscopic view on the market it is mainly manifested in several categories of market vision, such as the noise in the process of existing heterogeneous volatility in the stock price and transaction phenomenon and so on. Many of the existing study market phenomenons has close ties with investors trade behavior, this paper puts forward the HPIN as a branch of investor behavior research also has a close connection with various market phenomenon. By regression analysis and using methods such as "big data" opinion mining market information,we found that "herd" do exist in the Chinese market, and relationship between HPIN and market opinion also exist.At the micro level, between half-informed traders and uninformed traders exist games, this game of trading behavior under the traditional assumption is "irrational", the stock market is the direct reflection of its cause, and heterogeneous phenomenon of volatility and noise are two of them. In this paper, by establishing the model with heterogeneous volatility of stock prices, we use the GMM estimation method, the study found significant in China’s stock market volatility are heterogeneous and have not been effective pricing, shows that this kind of volatility is derived from different traders trading behavior rather than risk factors. The traditional understanding of noise is often as a white noise with zero mean, don’t participate in pricing. And this paper we use the Gamma-OU process description leap noise process to join the stock price path, design and using the method of Monte Carlo(MC) simulation with bouncing noise of stock option pricing process, found that the noise impact on asset pricing cannot be ignored.