Forecasting China's stock market volatility with shrinkage method: Can Adaptive Lasso select stronger predictors from numerous predictors?
本文用自适应Lasso方法从大量预测因子中筛选变量来预测中国股市已实现波动率，发现该方法在样本外预测中优于其他模型，尤其在低波动期表现更好。
Abstract This paper employs a shrinkage method named Adaptive Lasso (ALasso) to predict the realized volatility (RV) of the China's stock market with numerous predictors. We observed from the out‐of‐sample predictions that the ALasso model exhibits better predictive power than its competitors, implying that the ALasso method can select stronger predictors in the forecasting process than competing models. In addition, the predictability of ALasso method is better in low volatility periods than in high volatility periods. Finally, several robustness check methods, including different forecasting windows, different low and high volatility division, and different volatility measures, supported our results.