我国货币供应对经济增长的传导机制研究——基于六种AR-MIDAS模型的分析

Research on the Transmission Mechanism of Money Supply to Economic Growth——The Analysis of Six AR-MIDAS Models

引入多种权重函数构建六种混频数据时间序列模型(AR-MIDAS),以此研究我国的货币供应对经济增长的作用机制、传递路径及预测能力,多角度综合分析了六种模型样本内预测表现及估计效果,实证结果表明:Beta-AR(1)-MIDAS模型、Beta Non-Zero-AR(1)-MIDAS模型及Almon-AR(1)-MIDAS模型模拟结果及预测效果较优,特别是当滞后阶数变动34阶时,Almon-AR(1)-MIDAS模型表现最佳。短期内,8个月后,货币供应显著影响经济增长,扩张的货币政策能够拉动经济增长;长期内,特别是26个月直至34个月,货币供应对经济增长再次发挥出促进作用,其传导机制呈现先降后增的倒S曲线形路径。

In this paper, six AR-MIDAS models are constructed by the variety of weighing functions, this paper studies the effecting mechanism, transmission path and forecasting ability of money supply to economic growth in China, comprehensively analysises predictive performance in sample and estimation results of six AR-MIDAS models, The empirical results show that: simulation results and forecasting effect of Beta-AR( 1)-MIDAS model, Beta Non-Zero-AR(1)-MIDAS model and Almon-AR( 1)-MIDAS model is excellent, especially when the order of the lag is varied by 34 steps, the Almon-AR(1)-MIDAS model is best. In the short-term, eight months later, the money supply significantly has affect on economic growth, expansionary monetary policy can stimulate economic growth, in the long-term, especially from 26 months to 34 months, the money supply once again plays a promoting effect on economic growth which is conduction mechanism presents the first droping, after increasing of S curved path.