跳跃条件下中国证券市场资产价格行为研究

发布时间：2018-05-24 13:52

本文选题：资产价格跳跃 + 市场微观结构　；参考：《天津大学》2012年博士论文

【摘要】：随着资产价格行为研究的精细化和微观化，资产价格跳跃行为已成为市场微观结构等金融研究领域的重点和难点问题之一。跳跃行为的研究深刻揭示了金融市场资产价格发现过程的微观机理，直接影响资产价格波动率的估计和预测，继而促使传统的金融理论和计量方法必须进行深刻调整，并构成资产动态配置、风险管理和金融衍生产品定价等诸多金融实践领域的核心环节。在中国证券市场跳跃频发的背景下，本文基于证券市场微观结构理论视角，沿用“已实现”波动理论的主流研究框架，对资产价格跳跃行为进行了系统全面的理论和实证研究。具体内容如下： 1、跳跃行为的非参数辨识方法研究。基于BNS跳跃辨识理论框架，系统刻画了中国证券市场跳跃行为分布特征，包括对跳跃频度、久期、强度、方向、时刻分布、指数关联性、波动贡献等问题的完整讨论，，初步研究了剥离跳跃性收益的日间收益率分布特征。得到结论：中国证券市场跳跃频发，个股跳跃行为与市场整体状态关联紧密，不同方向的跳跃行为对日间收益率分布的作用机制存在显著的差异，剥离跳跃性收益的日间收益率序列的平稳性得到了明显加强。 2、基于信息冲击和流动性冲击的跳跃行为引发机制研究。首先对跳跃前后非对称信息与流动性分布特征进行了深入的刻画，在此基础上，构建了结合信息冲击和流动性冲击的Probit模型，对不同类型股票跳跃行为的引发机制进行研究。实证结果表明，单一因素无法对跳跃行为产生的原因进行完整诠释，并且对于不同类型股票，信息与流动性冲击的作用模式也是不尽相同的。 3、基于跳跃行为的波动率预测研究。首先在传统的HAR RV模型的基础上发展了具有方向性变差以及杠杆效应特征的HAR RS Leverage_(RV)模型，研究其对预期波动的作用机制。进一步，基于“已实现”半方差理论定义了方向性跳跃变差的概念，发展并构造了HAR-RV-C△J-Leverage_(C,J)模型，分析方向性跳跃变差与连续性变差以及杠杆效应因子对预期波动的作用机理。实证结果表明，HAR-RS-Leverage_(RV)模型适用于长期波动的预测，HAR-RV-C△J-Leverage_(C,J)模型对短期波动具有更好的预测效果。 4、基于有效波动率的市场一般性风险测度研究。首先依据鲁棒跳跃波动率medRV估计量，给出了有效波动率的定义并构建了有效波动率预测模型ARFIMA medR。V蒙特卡洛模拟实验得出结论：有效波动率估计量medRV能较好的鲁棒跳跃行为，能显著提高对波动率预测的准确程度。在此基础上进一步构造了市场一般性风险测度VaR medRV，实证研究表明：能有效摒除市场跳跃风险因子，对市场一般性风险进行有效测度。 5、证券市场跳跃性风险问题研究。首先对“已实现”贝塔系数β、“已实现”离散贝塔系数β~d、“已实现”连续贝塔系数β~c的分布特征及其内在关系展开了理论和实证研究，以此构成对市场跳跃风险系统性成分研究的技术基础；进一步，从资产收益的不确定性角度构造了市场日内跳跃风险测度，对中国证券市场不同类型资产在不同市场状态下的跳跃风险分布特征进行刻画，揭示了中国证券市场跳跃风险存在的系统性问题，并从价格发现与投资者行为角度对实证结果的市场根源进行了深入讨论。
[Abstract]:With the refinement and microrefinement of the research on asset price behavior, asset price jumping has become one of the key and difficult problems in the field of financial research, such as market microstructures. The study of jumping behavior profoundly reveals the microscopic mechanism of the asset price discovery process in the financial market, and directly affects the estimation and prediction of the volatility of asset prices. Then the traditional financial theory and measurement methods must be deeply adjusted, and the core links in many financial practice fields, such as the dynamic allocation of assets, risk management and the pricing of financial derivatives, have been used in the context of the securities market micro structure theory. The mainstream research framework of wave theory conducts a systematic and comprehensive theoretical and Empirical Study of asset price jump behavior.
1, study on the nonparametric identification method of jumping behavior. Based on the BNS jump identification theory framework, this paper systematically depicts the distribution characteristics of jumping behavior in China's securities market, including the complete discussion of jumping frequency, duration, intensity, direction, time distribution, exponential correlation, volatility contribution and so on, and preliminarily studies the daytime income of jumping income. It is concluded that the jump behavior of China's securities market is frequent, the jump behavior of the stock market is closely related to the overall state of the market. There is a significant difference in the mechanism of the interaction between different directions of jumping behavior on the distribution of daytime yield, and the smoothness of the daily yield sequence of peeling jumping income has been obviously strengthened.
2, the mechanism of jumping behavior based on the impact of information shock and liquidity shock is studied. First, the asymmetric information and the characteristics of the liquidity distribution before and after jumping are deeply depicted. On this basis, a Probit model combining information shock and liquidity shock is constructed, and the trigger mechanism of the jump behavior of different types of stocks is studied. The results show that the single factor can not complete the interpretation of the causes of jumping, and the modes of action of information and liquidity impact are not the same for different types of stocks.
3, the study of Volatility Prediction Based on jumping behavior. Firstly, based on the traditional HAR RV model, the HAR RS Leverage_ (RV) model with directional variation and leveraging effect is developed to study the mechanism of its effect on expected fluctuation. The HAR-RV-C Delta J-Leverage_ (C, J) model is developed and constructed to analyze the mechanism of the variation of directional jump and continuity and the effect of leveraged effect factors on expected fluctuation. The empirical results show that the HAR-RS-Leverage_ (RV) model is suitable for the prediction of long term fluctuations, and the HAR-RV-C Delta J-Leverage_ (C, J) model has a better prediction for short-term fluctuations. Effect.
4, based on the effective volatility of the market general risk measurement research. First, based on the robust jump volatility medRV estimates, the effective volatility is defined and the effective Volatility Prediction Model ARFIMA medR.V Monte Carlo simulation experiment results are concluded that the effective volatility estimation medRV can be better robust jumping behavior, can be shown to be significant. In order to improve the accuracy of Volatility Prediction, a general risk measure VaR medRV is further constructed.
It can effectively eliminate market jumping risk factors and effectively measure the general market risks.
5, the study of the jump risk in the securities market. First, the theoretical and empirical research on the "realized" beta coefficient beta, the "realized" beta coefficient beta ~d, the distribution characteristics of the "realized" beta coefficient beta ~c and its intrinsic relationship are carried out to form the technical basis for the systematic study of the market jump risk. Step, the market jump risk measure is constructed from the uncertainty of asset income, and the leaping risk distribution characteristics of different types of assets in different market states are depicted. The systematic problems of jumping risk in China's securities market are revealed, and the empirical analysis is made from the perspective of price discovery and investor behavior. The market roots of the results were discussed in depth.
【学位授予单位】：天津大学
【学位级别】：博士
【学位授予年份】：2012
【分类号】：F224;F832.51

【参考文献】