Empirical Analysis Of Security Markets Based On High-Frequency Data

As the high frequency finance data has a wealth of market information and is a new field of financial research, in recent years, the study of financial volatility is hot in financial field. Since realized volatility brought by Bollerslev etc., many researchers work for more in-depth study on high frequency finance data. For example, Neil Shephard etc. have been proposed to Realized Bipower Variation and Zhang Shi-ying have been proposed to Weighted Realized Variation in considering the calendar effect, but study are few on Weighted Realized Bipower Variation. So far study for transmission characteristics of the international security markets, the researchers applied the low frequency finance data to study. At present in the international security markets research, researchers are using the low frequency data, so it has information lose and securities markets trading is a coutinuous market price, the application of low frequency finance data is a big flaw to study for transmission characteristics of the international security markets. As the financial high frequency data has more information to the low frequency data, this innovative use of high frequency data research the transmission of international securities markets. The main achievements and the key point are follows:Make comparison between Realized Bipower Variation and Realized Volatility from the definition from. Result is in the application of Realized Bipower Variation is more robust and effective. Weighted Realized Bipower Variation is given based on the consideration of calendar effect, and prove its effectiveness and no bias.At present, using co-persistence in RV-VAR model for empirical research is few at home and abroad. In this paper, China,Hong Kong,Japan,the UK,the U.S. security markets index of 50,000 high frequency data use the model co-persistence by empirical analysis. Based on wavelet neural network theory, the model of non-linear co-persistence is set up to analyze the transmission of international security markets.Eliminate the structural industry factors of security and the impact of development of industry, empirical analyze in the specific in-depth study of high frequency data transmission between international security markets industry index. Base on Realized Volatility model, the potential volatility in the time series change into time-series volatility can be observed, can be modeled using conventional modeling techniques. This article uses industry and share price to study, apply WRBV to change into time-series volatility can be observed and uses Ganger causality test and variance decomposition to obtain better results.For the security market index and industry index is compiled by different enterprises, study the China National Petroleum share price,the China Life share price,the Aluminum Corporation Of China share price list in the Shanghai Stock Exchange, the Hong Kong Stock Exchange and the New York Stock Exchange to eliminate the influence of corporate, can be conducted only reflect the characteristics of the international security markets. Because enterprises have different share price in the different security markets only for security market factor.