中国股票市场跳跃现象研究

发布时间：2018-01-17 03:16

本文关键词：中国股票市场跳跃现象研究　出处：《天津大学》2016年硕士论文　论文类型：学位论文

【摘要】：对于股票收益率时间序列波动性中的跳跃现象的研究是近年来的研究热点之一,这主要是由于跳跃现象在金融经济学各领域的研究中均普遍存在,具有重要研究意义。同时近几年来受益于高频数据的使用,该现象的实证研究得以进一步发展。大量实证表明,通过将跳跃成分加入到金融资产价格的建模过程中,模型对于收益率,波动性等指标的拟合程度大幅提升;同时对于跳跃现象本身的研究也很好的揭示了收益率序列的非连续性变动特征。这使得对于跳跃现象的研究在风险对冲,产品定价等多个方面中都起到了重要的作用。本文在市场微观结构理论角度下,采用近期的使用广泛的非参数检验方法BTL法为基础,对作为沪深两市“晴雨表”的沪深300指数进行研究,从跳跃产生原因以及跳跃在波动性预测以及风险预测三个方面分别进行了实证分析。本文主要工作如下:我们选取了中国股票市场中能够反映沪深两个市场的沪深300指数,以2007年1月4日到2013年12月30日七年的股指5分钟交易数据作为研究对象,首先采用BTL法对沪深300指数进性跳跃的识别与分离,并进行统计性描述;随后使用Logit模型针对中国股市与中国宏观经济具有紧密联系的特点对跳跃产生的原因,即跳跃与宏观信息发布之间的关系进行研究分析;之后对现有的HAR-RV模型进行优化,将跳跃引入其中作为一个重要解释变量,提高了模型对于已实现波动率的预测结果;最后通过使用HAR-RV-up CJ-dwCJ模型对风险值VaR进行预测,并与传统的GARCH模型的预测结果进行对比,揭示了股票市场中的非连续成分,即跳跃对于收益率波动及风险值预测的重要作用。
[Abstract]:The research on the jump phenomenon in stock return time series volatility is one of the hot topics in recent years, which is mainly due to the phenomenon of jump in financial economics in all fields of research. At the same time, the use of high-frequency data in recent years, the phenomenon of empirical research has been further developed. A large number of empirical evidence shows. By adding the jump component into the modeling process of financial asset price, the fitting degree of the model for the return, volatility and other indicators is greatly improved; At the same time, the study of jump itself also reveals the discontinuous change characteristics of yield series, which makes the study of jump phenomenon in the risk hedging. The pricing of products has played an important role in many aspects. In the view of market microstructure theory, BTL method, which is widely used recently, is adopted in this paper. As a "barometer" of Shanghai and Shenzhen stock markets, the index of Shanghai and Shenzhen 300 is studied. This paper makes an empirical analysis from the reasons of jumping and the volatility prediction and risk forecasting. The main work of this paper is as follows:. We choose the Shanghai and Shenzhen 300 index which can reflect the two markets in the Chinese stock market. Taking the five-minute trading data of the stock index from January 4th 2007 to December 30th 2013 as the research object, the paper uses BTL method to identify and separate the advance jump of Shanghai and Shenzhen 300 index. And statistical description; Then the Logit model is used to analyze the reason of the jump, that is, the relationship between the jump and the release of macro information, aiming at the characteristics of the close relation between Chinese stock market and China's macro economy. Then the existing HAR-RV model is optimized and the jump is introduced as an important explanatory variable to improve the prediction results of the realized volatility. Finally, the HAR-RV-up CJ-dwCJ model is used to predict the risk value VaR, and the results are compared with the traditional GARCH model. It is revealed that discontinuous components in stock market, that is, jump, play an important role in the prediction of return volatility and risk value.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F224;F832.51

【相似文献】