中国货币政策不确定性对宏观经济的影响——基于MIDAS-VAR-CFSVM模型

Monetary Policy Uncertainty and Its Impact on Macroeconomics in China—Based on MIDAS-VAR-CFSVM Model

受限于VAR-SVM模型的结构特点,数量型和价格型货币政策不确定性对宏观经济的影响难以同时进行分析,本文构建MIDAS-VAR-CFSVM模型来克服这一难点。在新框架下,基于混频数据可以同时给出不确定性指数与脉冲响应结果,规避了两步法存在的一些问题,由于该框架包含因子随机波动率的结构,因此也容易对高维数据进行建模分析。基于MIDAS-VAR-CFSVM模型,本文利用中国1996年第一季度至2020年第三季度的数据,分析中国货币政策不确定性对宏观经济的影响。可以发现不确定性指数在2002年左右达到峰值,2007年后波动加剧,2016年后逐渐降低但在COVID-19冲击下再度攀升。不确定性冲击在短期内使得货币供应量增加和利率下降,引发价格水平上升、消费下降并且抑制经济增长,但对投资的影响不明确;在中长期内,反而会使投资增加,同时拉动消费和经济增长,且消费脉冲响应符号由负转正的时点先于经济增长。

We propose a mixed data sampling VAR with common factor stochastic volatility in mean model(MIDAS-VAR-CFSVM)to measure the uncertainty of quantitative and pricebased on monetary policy and its impact on macroeconomics simultaneously,which cannot be achieved by the standard VAR-SVM.The uncertainty index and impulse response function are estimated endogenously based on the mixed frequency data in our novel framework,which avoids many shortcomings of the two-step method.Our framework can also be used to model the factor stochastic volatility structure of the high-dimensional data.Based on China’s quarterly data between 1996 and 2020,we find that the uncertainty index reached its peak in 2002 and its volatility increased after 2007 financial crisis.The index gradually decreased after 2016 but rose again due to the shock of COVID-19.Impulse response shows that the shock of monetary policy uncertainty increases the money supply,decreases the interest rate,pushes up the price level,decreases the consumption and economic growth in the short term.Moreover,the impact of uncertainty shock on investment is not significant in the short run.In the long term,shock of monetary policy uncertainty generates positive effects on investment,consumption and economic growth,and the sign of response of consumption turns to positivity ahead of that of economic growth.