论文部分内容阅读
在金融时间序列波动具有显著的长记忆性这一背景之下,研究了LMSV模型长记忆参数的估计问题。首先,分析了LMSV模型的相关性质;接着,根据LMSV模型和ARFIMA模型的良好对应关系,提出了估计LMSV模型长记忆参数的半参数方法;最后,基于股市数据,验证了波动半参数方法的有效性。
Under the background that the financial time series fluctuation has significant long memory, the long memory parameter estimation problem of LMSV model is studied. Firstly, the related properties of LMSV model are analyzed. Then, based on the good correspondence between LMSV model and ARFIMA model, a semiparametric method for estimating long memory parameters of LMSV model is proposed. Finally, based on the stock market data, the effective half-parameter method is validated Sex.