| Due to the financial products are different from the other entities of goods and services, it is generally agreed that the whole financial economy is a special branch of the economy. The core issue of finance is to examine the allocation of capital and asset efficiency. In a market economy, this configuration is adopted for the financial markets. In recent years, with the rapid development of the world economy, the financial status in the economy growing, more in-depth researchs are done on the financial markets. Microstructure theory of financial markets is the fastest growing emerging as an important branch, which has gradually developed into one of the most active research areas in the modern micro-finance and greatly deepened about the operation of financial markets, financial transactions of products, price formation of financial assets and financial risk management awareness and control.The U.S. "subprime crisis" gradually began to appear in the spring of 2006. It began to sweep through he world's major financial markets, such as the United States, the European Union and Japan in the August of 2007 and evolved into the global financial crisis in 2008. China's stock market has also been inmpacted by the "sub-loan crisis" and undertaked the severed test. However, we can not be stalled. Compared with developed countries, the depth of China's stock markets are still far from enough. The exiting financial and securities products can not meet the people's growing demand for wealth in the 30 years of reform and opening. Faced with this demand, China's gradual deepening of financial and securities markets is inevitable which will become a new growth point for the stock market. This put forward higher requirements on the deeply study on market microstructure in China stock market, effective supervision of market, building effective and efficient high-quality transparent market.The conception of market microstructure is divided into the narrow and broad sense. In a narrow sense, market microstructure emphasize the price discovery of securities, the formation process of the price and the operational mechanism (O'Hara, 1995). In the broad sense, market microstructure involves almost all components of a complete market, including technology, regulation, information, participants and financial instruments(Zhishu Yang, 2000). In general, the market microstructure theory studies the the common law of all the financial markets, which has a strong abstract and universality. The market microstructure research methodology and theoretical framework are applied for the various financial markets, including stock, bond and foreign exchange market. By systematically combing the literature of market microstructure theory, the author find that although the foreign scholars describe the meaning of market microstructure from different point of view, the basic content is the same. Market microstructure reflects all aspects of the market itself, the features guides the action of investors, which determine the liquidity, transparency, the effectiveness of price, volatility and transaction costs in the market.After nearly two decades of rapid development, China's stock market has become not only the Asia's third-largest stock market, but also one of the largest emerging markets which adopts a limit order-driven trading mechanism. As an emerging securities market developed from transition economy, China's stock market has an unique market structure and constraint conditions. Policy market, civil market, game market, the shocked market and the transition market are still the main features. At the same time, we should also see that the current China's stock market has many features, suah as a high degree of segmentation of companies, the low concentration of the size, high exchange rate and so on. How to regulate, manage and develop the stock market is a priority. In the macro-control of reform on securities market, it is important for the organizers, regulators and participants in China stock market to enhance its efficiency.The overall structure of the thesis is as follows.Chapter 1 is an introduction. We introduced the research motivation and significance, summarized the major research methods, research contents, research tools, data sources and innovations.Chapter 2 is an overview of China's stock market microstructure. We overviewed on China's stock markets, the main research objects and contents of market microstructure theory. We compared the the microstructure between the New York Stock Exchange and Shanghai Stock Exchange, summed up the micro-structure characteristics of China stock market, which had laid a solid foundation in the theory and constraints for the further theoretical and empirical analysis of the market.Chapter 3 is the high-frequency characteristic variables of market microstructure in China stock market. This chapter reviewed the high-frequency financial study, put forward the concept of market microstructure feature variables for the first time, discussed definition and calculation method of a market microstructure characteristic variable from the perspective of the marked point process laying the modeling basis for the thesis. We presented an overview of high-frequency financial database, and further gived the sample selection and pre-processing data for the study of China stock market about the high-frequency data, which provided the raw material for the empirical analysis and laid a data foundation for the thesis.Chapter 4 is the statistical characteristics of the high-frequency variables of market microstructure in China stock market. Based on intraday transaction data, this chapter comprehensively studied the descriptive statisticscs and distribution laws of the market microstructure characteristic variables for the first time. We gived the descriptive statistics and unconditional distribution of the variables, such as stock returns(price changes), trading volume, bid-ask spread, the transaction duration, the price duration and volume duration which are closely related to trading activities. In particular, in-depth and detailed research on financial durations of financial transactions in the aspect of the tading duration, price duration and volume duration is an innovation in this chapter.Chapter 5 is the research on intraday pattern of characteristic variables of market microstructure in China stock market. Based on the intraday data, we comprehensively studied the intraday pattern of characteristic variables of market microstructure in China's Shanghai and Shenzhen A, B-share market. We used the parametric and nonparametric methods to test the intraday dynamic changing pattern of stock returns (price changes), volatility, trading volume, bid-ask spread and financial duration which are closely related to trading activities and compared them with the conclusion of the mature markets in foreign countries. Finally, we used the linear and nonlinear Granger causality test to reveal the non-symmetry of the transmission of information. Combing with the overnight accumulation of information transmission and information asymmetry, we explained the reasons of the intraday patterns and put forward some suggestion to the design of the trading mechanism and regulation of information.Chapter 6 is the quantitative research on the financial duration in China stock market. This chapter first systematicly reviewed the theoretical model and the economic and financial applications of the ACD model. In section 6.1, we summed up the market microstructure meaning of the financial durations and made five key assumptions. In section 6.2, we introduced the theoretical basis of duration, properties of ACD model and the corresponding estimation method. A Monte Carlo simulation method was used to evaluate maximum likelihood estimation and nonparametric estimation method. The simulation results show that the misspecification on distribution of random error leads to a serious failure in maximum likelihood estimation, nonparametric estimation method can be used to select the appropriate distribution function of random error when the mean function form and the distribution of random error are unknown. In section 6.3, we summed up the expansion and development of ACD mocel in the five aspects. In section 6.4, we summarized and evaluated current three important tests of ACD model, put forward a test method based on NP estimates of the ACD model, which is easy to use and suitable for nested and non-nested test of the model. In section 6.5, first of all, we applied Granger causality test to explore the information tansimission relationship between the characteristic variable of market microstructure and trading duration and their forecasting capability to the duration. Then we built a multi-linear ACD (MLACD) and semi-parametric single index ACD (InACD) model and its estimated algorithms to test the five assumptions on market microstructure theory in China stock market. Finally, we discussed the intraday risk measurement based on ACD model, Irregularly Spaced Intraday Value at Risk(ISIVaR) model was used to estimate the conditional volatility of intraday transaction and give the VaR forecast and the backtestingof real-time trading price changes, which make up for the current research gaps in this regard.Chapter 7 is panel data analysis of price impact factors in China stock market. First of all, based on the linear and nonlinear Granger causality test, we verified the heterogeneous properties of the information transmission of stock market microstructure variables. We then discussed Granger causal relationship between the market microstructure characteristic variables and their transmission mechanisms on the basis of fixed effect panel data model under the heterogenous Causality (HNC) hypothesis testing in China stock market, which provides an information support for the price panel data models. In section 7.3, based on the intraday 15-minute high-frequency panel data, we established the fixed effect panel data model of bid-ask spread variable with changing intercept to study the impact of the weekday pattern, intraday pattern and other some factors on bid-ask spread in China's Shanghai and Shenzhen A, B-share markets. In section7.4, based on the intraday 15-minute high-frequency panel data, we established the fixed effect panel data model of price variable with changing cofficient to study the impact of the status of the characteristics of limit order and market microstructure characteristic variables on price in China's Shanghai A, B-share markets. In section 7.5, based on the intraday 15-minute high-frequency panel data, we established the fixed effect panel data model of the up and down of price with changing intercept to study the impact of the the status of the characteristics of limit order and market microstructure characteristic variables on the up and down of price in China's Shanghai A, B-share markets. We also tested the predicted ability of the the status of the characteristics of limit order and market microstructure characteristic variables on the up and down of price.Chapter 8 is summary and outlook. We summarized the research contents and main conclusions in the whole thesis, and then put forward the prospect of future research.In conclusion, based on previous studies, the innovation is as follows.1. Innovative ideas on research. In this thesis, we summed up and proposed the concept of the characteristic variables of market microstructure. By exploring into the properties of trading activities of the duration process and the marked point process, we want to reveal the "black box" mystery from the limit order in the trading process, research on representive market microstructure characteristic variables deeply, piont out a feasible path for further study on the market microstructure theory. Three aspects of "linear and non-linear model used at the same time", "parameters and nonparametric time series analysis accrossed," "intraday time series data and panel data complemented each other" characteristics of the study when we analyzes the intraday pattern, the financial durations and price discovery in the thesis.2. In Chapter 6, a test of ACD model is proposed based on the NP estimation, a multivariate linear ACD (MLACD) and a semi-parametric single-index ACD (InACD) model and its estimate algorithm are built to test the five assumptions on market microstructure theory in China stock market. We also discussed the intraday risk measurement based on ACD model, Irregularly Spaced Intraday Value at Risk (ISIVaR) model is used to estimate the conditional volatility of intraday transaction and give the VaR forecast of real-time trading price changes, which make up for the current research gaps in this regard.3. In chapter 7, from the point of view of panel data, we study the impact factors of the bid-ask spread in China's Shanghai and Shenzhen A, B-share market, the impact factors of price (changes) and the up and down of price in China's Shanghai A, B-share market based on the panel data unit root test, Granger causality test and cointegration test for the first time, which reflect the mechanism that a variety of transaction information influce the price at and some extent and reveal the price discovery mechanism in the course of transaction.4. In chapter 5, from the point of view of intraday high-frequency data, the parametric and nonparametric methods are used to test the intraday dynamic changing pattern of stock returns(price changes), volatility, trading volume, bid-ask spread and financial duration which is closely related to trading activities. The results are compared with the conclusion of the mature markets in foreign countries. The linear and nonlinear Granger causality test are used to reveal the non-symmetry of the transmission of information. Combing with the overnight accumulation of information transmission and information asymmetry, we explained the reasons of the intraday patterns and put forward some suggestion to the design of the trading mechanism and regulation of information.As a result of market microstructure research involving the areas of more extensive, any attempt to cover all areas of theoretical and empirical research will be superficial and unrealistic. In the thesis, from the view of characteristic variables of market microstructure, by exploring into the properties of trading activities of the duration process and the marked point process, we analyzes the intraday pattern, the financial durations and price discovery in the thesis. However, due to the profound theory of the market microstructure and the trading activities are very complex, we should deeply research on intraday risk measurement and and price discovery mechanism, on how to use the quantitative conclusions and evaluate the policy. In addition, because of limiting of writing time, as well as restrictions on sources of data, the thesis use the dada in 2006 and 2007 when the stock market is in bull market. More extensive research on the bear market should be done to make up for the shortcomings of the thesis, the conclusions will be more robust. However, the research perspective, research methodology and research framework of the thesis broaden the quantitative study and empirical analysis of the market microstructure theory. We point out a feasible characteristic path of research on quantitative characteristics of the market microstructure, which is useful for further study.
|
PDF Full Text Request