摘要
经典的波动率模型(GARCH等)是收益率为基础的模型,利用的是收盘价信息,忽略了价格波动的日内信息,这将导致信息与效率的损失。为了弥补这一缺陷并获得满意的波动率预测效果,本文引入并扩展了基于价格极差的自回归波动率模型。实证研究表明新模型能够有效刻画波动率的动态变化规律,其预测效果一致性地优于经典的GARCH模型。同时,我们的研究还证实了在波动率模型中加入收益率的滞后项能够提高模型的解释能力,并且存在明显的"杠杆效应"。
The classical volatility models, such as GARCH, are return-based models, which are constructed with the data of closing prices. It might neglect the important intraday information of the price movement, and will lead to loss of information and efficiency. This study introduces and extends the rangebased autoregressive volatility model to make up for these weaknesses and obtain satisfactory volatility predicting performance. The results consistently show that the new model successfully captures the dynamics of the volatility and gains good performance relative to GARCH model. Furthermore, we find that the inclusion of the lagged return can significantly improve the forecasting ability of the volatility model, and the leverage effect does exist in volatility.
出处
《中国管理科学》
CSSCI
北大核心
2009年第6期1-8,共8页
Chinese Journal of Management Science
基金
国家自然科学基金委员会优秀创新研究群体基金(70221001)
教育部人文社会科学研究项目
湖南师范大学社会科学青年学术骨干培养计划(基金)
关键词
波动率建模
价格极差
日内信息
预测绩效
volatility model
price range
intraday information
forecasting performance
