Jump Behavior Research On Financial Assets Prices

The behaviors of stock price jump has been one of the most important research fields in the Market Microstructure theory. Price jump influence the accuracy of volatility estimating and forecasting distinctly, in turn severely affect asset allocation and risk management. And it also precipitates a profound readjustments of the conventional econometric models and quantitative methods. Consequently, with the research on asset price more refined and microscopically, the jump issue must be addressed, which is not only profound academic significance, but also positive practical significance. And the analysis of jump characteristics of China Stock Market, as well as revealing deeply the essence of price-discovery process, are of great practical significance for perfecting market trading mechanisms and maximum market effectiveness.Under Market Microstructure angle of view, this dissertation presents a comprehensive study to explore volatility estimation under jumps, generalized behavior of jumps indentification, describing jump tail distribution, pricing jump risk followed the generalized p-power variation framework. The detailed contents are as below:1. Research on Realized Kernel Volatility Estimation method under jump. Study the accuracy of RK and RV estimation methods on the basis of simulated path using Monte Carlo method, the result is that the disturbance of the discontinuous factors could be eliminated by RK, which makes the estimates more similar to IV. Moreover, the RK-AFRIMA model is constructed to estimate and forecast volatility in China Stock Markets using high-frequency data on the basis of modified fraction order difference algorithm, and get conclusions that RK has better applicability for China stock market and better prediction results than RV.2. Research on jump identification with Nonparametric Method. Consider the identification of generalized jump behavior with TMPV method compared with BNS method. Obviously, BNS is just applied to lower high frequency data and used to identify finite activity jump. However TMPV can depict generalized jump based on ultra-high frequency data. Then corrects the threshold time-varying problem in new lines of investigation into generalized jump behavior with TMPV method and discusses the behavioral genotype's sampling frequency dependency in generalized jump behavior research, and provide causes of this phenomenon on the empirical research results.3. Research on jumps behavior Price-Discovery based on information structure. Analyze the dynamics process of jump phenomenon formation incorporated with information shocks, presents evidence that jumps serve as a dramatic form of price discovery fully involved into information and make no significant effect on future return, so haven't pricing ability. Furthermore, the Information Distribution Measurement Model is constructed, and then characterizes the levels of Information Distribution Imbalance with different types of securities in China Stock Market and presents a discussion on the pricing problem. The results demonstrate that jump information distribution has explanatory power for the overall market as pricing factor, and negative effects to return except for large-cap, and the ultimate reasons are the highly speculation in Chinese stock exchange and no compensations for the information risk.4. Research on Jump Risk pricing in China Stork Market. Analyze the statistical characters after detaching jumps from stock price and consider the features of jump behavior with different types of stocks in China Stock Market based on finite activity jumps driven by information. Furthemore constructs the Jump Risk Pricing Model and decompose the market risk effectively,and then a systematic study on risk features , risk ratios and effect among them is done .5. Research on distribution of return jump tail. Construct Generalized Pareto Distribution Model for return jump tail based on finite activity jump driven by information, and consider return jump tail distribution characteristics in China Stock Market. Furthermore, establishes measures of extreme value VaR and ES, and portrays risk characteristics of return jump tail in China Stock Market.