The Study Of Relationship Between Volume-Price Based On High-Frequency Data In Treasury Bond Futures Of China

This paper will study the relationship between the volume and price of Treasury Bond Futures based on high frequency data. Divide the price into bond futures yields (including Rate of return. Absolute Return), GK volatility, and divide the volume into Trading Volume and Position Volume. By Granger causality test, analyzing the relationship between causality yields, absolute return and trading volume, position volume of Treasury Bond Futures. Then according to the results of Granger causality test, the paper will build VAR model using three variables:the absolute return, the Trading volume and the Position volume of Treasury Bond Futures. Then this paper decompose variance and analyze impulse response, observing the incidence of the three variables and reaction to the sudden shocks. And the paper analyzes the microstructure of bond futures.Then analyzing the correlation of price volatility, trading volume and position volume, three regression models are built. Using the HAC regression estimation models, this paper will observe whether Trading Volume and Position Volume have a significant effect on the price volatility of Treasury Bond Futures. First, this paper divides the Trading Volume and Position Volume into predictable and unpredictable types by using ARMA model. The previous information variable is generated by investors' adjusting positions or liquidity demand, it's a predictable variable. The unpredictable variable is generated due to new information reaching the market, it cannot be explained by the sequence of historical amount. Then we introduction dummy variables and divide the variables related to new information variable into positive and negative impact. The paper analyzes their impact on the bond futures price volatility, thus analyzing bonds futures market way of transmitting information and efficiency in an overall.Through the empirical study, this paper has found that:absolute returns and Trading Volume has a two-way Granger causality. There is a one-way causal relationship between Position Volume and absolute return; There is no relationship between rate of return with Trading Volume and Position Volume. This general conclusion is different from the stock market, which indicates that short selling has a significant impact on the market microstructure. Through building VAR model by these three variables, the paper found that the lag phase Trading Volume has a significant positive impact on the absolute rate of return and the lag of Position Volume has a significant negative impact on the absolute return. And through the variance decomposition and impulse response analyzing, the paper found that the spread of bonds futures market information does not reach complete mixing distribution state.Through building regression modeling by price GK volatility, Trading Volume and Position Volume, which found that the Trading Volume has a significant positive impact on price GK volatility and the Position Volume has a significant negative impact on price GK volatility. This conclusion is similar to the VAR model. It explains that Trading Volume and Position Volume do have a significant effect on price GK volatility of Treasury Bond Futures.In this paper, the structure and contents are as follows:First, the paper illustrates the research's background and significance, stresses the main purpose of studying the Relationship Between Volume-Price Based on High-Frequency Data in Treasury Bond Futures of China. Then this paper describes the research methods, and finally reveals the innovations herein and shortcomings.The second chapter is literature review, mainly describing the correlation relationship between return and Trading Volume, Position Volume, and the correlation relationship between price GK volatility and Trading Volume, Position Volume. At last, stresses the research priorities and innovation of the paper by the method used in the relevant literature and empirical conclusions.The third chapter is about the Treasury Bond Futures market volume-price relationship and microscopic structures, analyzing the microstructure of bond futures market from the perspective of the relationship between volume and price respectively, describing the relationship of volume and price information transfer theory, reintroducing the effectiveness of Treasury futures market.The fourth chapter is the empirical part, first, processes the data and analyzes the descriptive statistical for the relevant indicators. Then through Granger causality test, observes causal relationship between trading volume and position volume, building VAR model by absolute return, volume and open interest. Variance Decomposition and Impulse Response analysis, quantitative analysis of the specific effects between them, so as to reveal the microstructure of the bond futures market. Finally, establish a three-layer progressive regression models by bond futures volatility, trading volume and position volume, and analyzes concretely the impact of their relationship.The fifth chapter is the conclusion and prospect. It's a summary of the full text, and the prospect of their own research shortcomings.Comparing with previous studies, this article has innovations in the following areas:(1) Study the correlation between volume-price in Treasury Bond Futures of China, refine the price of bond futures into the rate of return, absolute return and price volatility, and extends to the volume and open interest. So the object of study is comprehensive. It not only expands the research in Treasury Bond Futures of China, but also provide some reference for the study of bonds futures volume and volume-price relationships.(2) Employ the high frequency data to study the relationship between Volume-Price of Treasury bonds Futures in China.1 minutes and 5 minutes information is superior to the day data which is lack of relevant information. The high frequency data will make the record of information more perfect, and ensure the integrity of information.(3) The GK volatility is better than traditional volatility in high frequency data, and more efficient.