基于不同频率数据的波动率模型的研究

发布时间：2018-03-03 03:29

本文选题：波动率　切入点：ARFIMA模型　出处：《湘潭大学》2014年硕士论文　论文类型：学位论文

【摘要】：股票市场形势复杂,变化多端,对其波动率的研究已经成为该领域的核心内容之一。我国沪深股票市场为新兴金融市场,自实施涨跌幅限制以来已趋于稳定,但市场体制还需进一步地完善。波动性是衡量股市质量和效率的重要指标,一定幅度的波动有利于活跃股市并使其持续发展,过于频繁且剧烈的波动反而会适得其反。因此,波动率的研究对于我国股市的规范以及投资者的理性判断具有重要意义。本文以上海和深圳股市最具代表性的上证指数和深圳综指为研究对象,采用最新历史数据来分析我国近年股市波动情况。并选取了每只股票的日数据、5分钟数据和1分钟数据,对其建立两种均值模型和六种GARCH方差模型进行波动率的研究,得到两个结论： 1、最适合研究近年我国沪深股市收益波动率的均值模型为ARFIMA模型,方差模型为EGARCH模型。 2、不同频率的收益序列,其最优波动率模型不同,且频率越高,波动率模型的拟合效果和预测效果越好。本文1分钟数据的波动率模型的研究结果最理想。原因有三： 1、随着我国股票市场体制的完善,逐渐趋于稳定,股市收益时间序列已经具备长记忆性。 2、ARFIMA-EGARCH模型综合考虑了我国股市收益序列的自回归异方差性、杠杆效应、长记忆性和风险对收益的影响。 3、数据频率越高,对股市信息反映的更全面,更利于波动率模型的拟合和预测。
[Abstract]:The situation of stock market is complex and changeable. The research on volatility has become one of the core contents in this field. China's stock market in Shanghai and Shenzhen is an emerging financial market, which has become stable since the implementation of the limit of price and decline. But the market system needs to be further improved. Volatility is an important index to measure the quality and efficiency of the stock market. A certain range of fluctuations is conducive to the active and sustainable development of the stock market. The study of volatility is of great significance to the regulation of Chinese stock market and the rational judgment of investors. This paper takes the Shanghai and Shenzhen Composite Index, which is the most representative stock market in Shanghai and Shenzhen, as the research object. The latest historical data are used to analyze the stock market volatility in recent years in China. The daily data of 5 minutes and 1 minute of each stock are selected to study the volatility of two mean models and six GARCH variance models. Two conclusions are drawn:. 1. The ARFIMA model and the EGARCH model are the most suitable models to study the volatility of Shanghai and Shenzhen stock markets in recent years. 2. The optimal volatility model is different for different frequency return sequences, and the higher the frequency, the better the fitting effect and prediction effect of volatility model. There are three reasons:. 1. With the perfection of the stock market system, the stock market income time series has long memory. 2ARFIMA-EGARCH model synthetically considers the influence of autoregressive heteroscedasticity, leverage effect, long memory and risk on the return of stock market in China. 3. The higher the data frequency is, the more comprehensive the stock market information is, and the more favorable the volatility model is to fit and forecast.
【学位授予单位】：湘潭大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F830.91;F224

【参考文献】