期刊名称：Brazilian Journal of Probability and Statistics
印刷版ISSN：0103-0752
出版年度：2006
卷号：20
期号：2
页码：111-132
出版社：Brazilian Statistical Association
摘要：The idea of volatility is fundamental to precise definition ofrisk and, hence, its estimation (or prediction) is a very important task infinance applications. We present some ideas on non-parametric estimationof volatility function in di.usion models. A nonlinear wavelet estimate ofthe volatility function is proposed and its performance is compared to threekernel estimators in both simulated and real data. Simulation is developed foreight volatility shapes and some interesting, but not unexpected, results arepresented. Some issues such as online estimation and prediction, robustness tooversmoothing and performance under sudden changes in pattern of volatilityare also discussed
关键词：High frequency financial series; non-parametric curve esti-;mation; volatility estimation; wavelets
