股票收益率波动性的非参数核回归估计及对中国股市的实证分析

Nonparametric kernel regression estimation of volatility of stock return and its application to Chinese stock markets

非参数回归估计是研究非线性时间序列的一种有用工具.在混合相依样本的条件下,基于非参数核回归估计方法来研究收益率的波动性,利用改良的交叉核实函数选取光滑参数,并给出上海和深圳股市实证分析的一些有趣结果.

Nonparametric regression estimation is a useful technique for investigating nonlinear time series.Under mixing conditions,this paper gives kernel regression estimation of the volatility of the stock return by improving cross-validation function on choosing smooth parameters.Its application has been tried in Shanghai and Shenzhen stock markets in China,giving some interesting results.