Given that policy uncertainty shocks in the economic environment can exacerbate financial market volatility and pose financial risks, this paper utilizes a smooth transition version of the GARCH-MIDAS model to investigate the impact of different structural state changes in economic policy uncertainty (EPU) on stock market volatility. The extended model explains the nonlinear effects of the macro variables and the structural break changes in regime transitions. The empirical results confirm that the EPU indicators provide effective prediction information for stock volatility from the in-sample and out-of-sample analyses, which reveals that the smooth transition model provides an effective method for detecting the possible regime changes between stock volatility and macroeconomic uncertainty. Additionally, we further confirm that some category-specific EPU indicators also have strong smooth transition behaviour with respect to stock volatility. More important, our new model provides significant economic value to investors from a utility gain perspective. Overall, the institutional changes present in EPU play a nonnegligible and important role in stock market volatility. Accurate identification of the structural features of financial data helps investors deepen their understanding of the sources of stock market volatility.

鉴于经济环境中的政策不确定性冲击会加剧金融市场波动并带来金融风险，本文利用GARCH-MIDAS模型的平滑过渡版本来研究经济政策不确定性（EPU）的不同结构状态变化对股票的影响市场波动。扩展模型解释了宏观变量的非线性效应和制度转变中的结构性断裂变化。实证结果证实，EPU 指标从样本内和样本外分析中为股票波动提供了有效的预测信息，这表明平滑过渡模型为检测股票波动与宏观经济之间可能的制度变化提供了一种有效的方法不确定。此外，我们进一步证实，一些特定类别的 EPU 指标在股票波动方面也具有很强的平滑过渡行为。更重要的是，我们的新模型从效用收益的角度为投资者提供了显着的经济价值。总体而言，EPU 中存在的制度变化在股市波动中发挥着不可忽视的重要作用。准确识别金融数据的结构特征有助于投资者加深对股市波动来源的理解。