Research On Asset Price Behavior In China Stock Market Based On Price Jumps

With the research on asset price more fine and microscopically, the behaviors ofstock price jump has been one of the emphases and difficulty in the MarketMicrostructure theory and the research fields of finance. The research on price jumpdeeply reveals the microscopic mechanism of price discovery for financial market,directly influences the degree of precision in volatility estimating and forecastingdistinctly. It afterwards makes the traditional financial theories and econometricmethods must be profoundly adjusted, and constitutes the core link for the assetsdynamic configuration, risk management and financial derivatives pricing, and alsofor many other financial practice areas. Under the background of frequent jumps inChinese Securities market，this paper presents a comprehensive study on the behaviorof assets price jump from the perspective of financial Market Microstructure Theoryalong with realized volatility research framework. The main contents of thedissertation consist of five theoretical and empirical parts:1. Research on jumps identification with Nonparametric Method. Based on theframework of BNS theory, described the assets jump behaviors in China Stock Marketcomprehensively, including jump frequence, duration, intensity, direction, timedistribution, the relevance of market’s index and the proportion of jump in RealisedVolatility and so on. The results indicate jump behaviors happen frequently in ChinaStock Market, and jumps of individual share are closely related with the whole marketstate. Also take a preliminary study on the distribution of daily yield rates striped ofjumps, it turns out that the directions of jump have different effects of daily yieldsseries, and the ones without jump become smooth.2. Research on jump behaviors price discovery based on information shocks andliquidity shocks. First provide evidence that both information shocks and liquidityshocks have explanatory power in jump phenomenon formation, and then examine theasymmetric information and liquidity around jumps, and finally specify and estimate aProbit model to further explore the interactions between information and liquidityshocks for jumps of different types of stocks. The empirical results further confirmboth of the proposed factors have significant predictive power for jumps and theirmechanisms for different stocks are not the same. 3. Predictability of Volatility on the basis of jumps. DevelopHAR RS LeverageRVbased on the model HAR-RV, which decomposedrealized volatility into signed components and incorporated with the leverage effect ofrealized variance with an indicator for negative daily returns, to explore the effects ofthe signed realised volatility and its leverage for future volatility. And then formulatea new modelHAR-RV-C△J-Leverage_C,J containing signed jump variation aswell as leverage estimators of variations due to the continuous part and the jumps,which is used to investigate the role of them in future variance. It finds thatHAR-RS-Leverage_RV is more suitable for estimate long-term future volatility, andis of great value to predict short-term volatility.4. Measurement of Market General Risk Based on Effective Volatility. Constructa new measures of effective volatility by jump-robust volatility estimator medRV,and further establish a model for predicting effective volatility. And then test thevalidity and accuracy forecasting of the estimator with the simulationexperiment using Monte Carlo method. The simulation results validate that theestimator displays effective robustness to jumps as the effective volatilityand has better accuracy of volatility forecasting. Furthermore construct themeasurement of the market general risk VaR medRVbased on the research above,and the empirical evidence in China Stock Market indicate that the indicatorcan effectively reduce the extreme market risk factors and obtain theprecise measurement of the market general risk.5. Reasearch on Market Jump Risk. First part is about the theory and empiricalstudies of the distribution and correlations of realised betas, realised continuous betasand realised dis-continuous betas, which constitute the technology basis for theresearch on the systematic market jump risk. Further construct the measurement of themarket jump risk form the angle of the uncertainty of asset returns. And then describethe distribution characteristics of market jump risk of different types stocks underdifferent market conditions in China Stock Market. The empirical results reveals thesystematic characteristics of market jump risk in China Security Market and criticallydiscuss root causes of these results in the market from the point of price discovery andthe behavior of investor.