基于高频股票流动性的通货膨胀率预测研究  

Research on the Inflation Prediction Based on High-FrequencyStock Liquidity

作　　者：徐杨 吾俊达 XU Yang;WU Junda(School of Economics and Management,University of Chinese Academy of Sciences,Beijing 100190)

机构地区：[1]中国科学院大学经济与管理学院,北京100190

出　　处：《科技促进发展》2024年第4期369-376,共8页Science & Technology for Development

摘　　要：通货膨胀的准确预测对于投资者和政策制定者来说至关重要。经济的结构性转变给通货膨胀率的预测带来前所未有的挑战。本研究基于自回归混频数据抽样模型(AR-MIDAS)考察1996年6月至2023年12月中国A股市场流动性的日度信息对通货膨胀率短期预测的影响。结果表明,加入日度流动性指标后,AR-MIDAS模型对于通货膨胀率的预测性能具有显著提升。本研究创新性地将股票流动性加入通货膨胀预测模型,拓展利用金融市场信息预测通货膨胀的文献脉络,为认识金融市场与实体经济的关联提供新的视角,同时也为投资者和政策制定者提供决策参考。Accurate prediction of inflation is crucial for investors and policymakers.The structural transformation ofthe economy presents unprecedented challenges to the forecasting of inflation rates.This study examined the impact ofdaily information on liquidity in the Chinese A-share market from June 1996 to December 2023 on the short-termprediction of inflation rates,based on the Autoregressive Mixed Data Sampling model(AR-MIDAS).The results indicatedthat the inclusion of daily liquidity indicators significantly enhances the predictive performance of the AR-MIDAS modelfor inflation rates.Innovatively,this study incorporated stock liquidity into the inflation forecasting model,expanding theliterature on the use of financial market information to predict inflation.It provided a new perspective for understandingthe connection between financial markets and the real economy,and also offers decision-making references for investorsand policymakers.

关 键 词：中国通货膨胀率 混频数据模型 股票流动性 

分 类 号：F822.5[经济管理—财政学] F832.51