基于日度低频价格的波动率预测

Volatility forecasting based on daily frequency prices

利用作者提出的GARCH-X的框架,将以往文献中提出的各种基于金融资产的最高、最低、开盘和收盘等低频价格信息的波动率静态估计,统一地扩展成对波动率的动态预测模型.通过对上证指数近十几年数据的实证分析,并借助于对波动率的高频估计和预测评估的一些最新研究成果,本文揭示出合理地利用价格极差及开盘价的信息可以显著地提高对波动率及风险价值的预测能力.

Within the GARCH-X framework put forward by the authors, this paper considers several new vola- tility forecasting models based on daily high, low, opening and closing prices of financial assets. These models combine the GARCH modeling procedure and the results of volatility estimation in the early literature, and therefore extend the static estimators into the dynamic driving factors of volatility. Empirical results with the daily prices of the Composite Index of Shanghai stock market over the last decade reveal that the forecasting performances of these new models for volatility and Value-at-Risk are significantly better than the traditional GARCH model.