王阳， 雷钦礼^*

（  暨南大学经济学院，广州 510632；  ）

摘要： 本文以沪深300指数为例，将基于其5分钟高频数据的已实现波动率模型和基于其日收益数据的历史波动率模型这两类波动率模型的样本外日波动率的预测能力进行比较。结果发现，无论是样本内预测还是样本外预测，已实现波动率模型比传统的EGARCH模型有更好的预测能力；把已实现波动率作为解释变量加入传统的EGARCH模型的方差模型之后，有助于提高模型对数据的拟合能力和预测能力。
关键词： 已实现波动率；波动率模型；高频数据；波动率预测

Wang Yang， Lei Qinli^*

（  Economics School, Jinan University, Guangzhou 510632；  ）

Abstract： Taking the Shanghai and Shenzhen 300 Index as an example, the forecasting ability of the volatility model based on the 5-minute high-frequency data and the historical volatility model based on the daily return data Comparison. It is found that the realized volatility model has better prediction ability than the traditional EGARCH model both in intra-sample prediction and out-of-sample prediction. After adding the realized volatility as the explanatory variable to the variance model of the traditional EGARCH model, To improve the model's ability to fit the data and predictive ability.
Keywords： Realized volatility; volatility model; high-frequency data; volatility prediction

PDF全文下载：     初稿（4）

【收录情况】