摘要
本文是对证券市场波动率期限结构预测的一个尝试。本文采用上证A指在1994-12-21至2002-12-31期间内的日交易数据之收盘价作为研究对象,用AR、GARCH、GJR、GCOMP等模型预测出其在1997-7-2至2002-12-31期间的波动率期限结构,并将预测结果与实际结果进行比较分析。实证结果表明,尽管各种预测模型均存在一定缺陷,就短期(周或月)波动率的预测来说,GJR模型的预测效果最好,其次是GARCH模型,GCOMP模型表现最差,但就长期(季或年)波动率的预测而言,GCOMP模型表现较好。
This paper attempts to study the forecasting volatility term structure of security market. The data is based on the close price of Shanghai A-Shares from Dec.21st,1994 to Dec.29th,2002. We use AR, GARCH, GJR, GCOMP, models etc. to forecast, and get the forecasting volatility term structure of Shanghai A-Shares from July.2nd, 1997 to Dec.31 st,2002, and then compare them to the real volatility term. The physical result shows： When predicting short-term （week or month） ,volatility term structure, GJR model performs better than other models, GARCH model ranks second, and GCOMP model is the worst. However, GCOMP model does a good job when predicting long-term （ quarter or year） volatility term structure.
出处
《福建金融管理干部学院学报》
2007年第2期19-25,共7页
Journal of Fujian Institute of Financial Administrators
关键词
上证A指
预测
波动率期限结构
Shanghai A-Shares Index
forecasting
volatility term structure
