技术交易规则预测能力与收益率动态过程——基于Bootstrap方法的实证研究

Predictability of Technical Trading Rules and Return Dynamics: An Empirical Research Based on Bootstrap Method

本文对我国股票市场技术交易规则预测能力进行了实证检验,发现移动平均规则所产生的买入区间收益率更大而波动率却更小,卖出区间的收益率为负而波动率却更大。运用自举(Bootstrap)方法检验发现,四种常用的收益率线性模型均不能解释买卖出区间收益率与波动率所表现出的非对称现象,尤其无法解释卖出区间收益率为负的现象。为此,本文通过人工神经网络方法,将条件异方差结构引入到现有的收益率非线性模型,发现该模型能更好地解释买卖出区间收益率与波动率模式,表明收益率动态过程中存在非线性特征。

This paper investigates the predictability of moving average rules on China stock market. We find buy signals generate higher returns and less volatility, while returns following sell signals are negative and more volatile. Moreover, the bootstrapping results indicate that the asymmetrical patterns of return and volatility can not be explained by four popular linear models of returns. We then test the nonlinear dynamic process of returns. We introduce the conditional heteroskedasticity structure into the existing artificial neural network （ANN） model and find the revised ANN model can explain the predictability of returns and volatility better, which indicates some hidden nonlinear properties in returns dynamic process.