货币政策规则设定、外生冲击与中国宏观经济波动——基于动态随机一般均衡模型的分析

Setting of Monetary Policy Rules,Exogenous Impacts and China's Macroeconomic Fluctuations:An Analysis Based on the Dynamic Stochastic General Equilibrium Model

为了分析中国的宏观经济波动,可以用贝叶斯方法估计一个包含多种冲击和摩擦性因素的动态随机一般均衡模型。这类模型通常使用利率规则测度货币政策冲击,不过,货币数量规则可能更适合中国现实。在使用货币数量规则这种新的设定下,历史方差分解的结果表明,1992—2016年期间,产出增长率波动的主要驱动因素是货币供给冲击、投资冲击和外生需求冲击,而其中的两次剧烈波动则需要用其他冲击来解释;通货膨胀率波动的主要驱动因素是货币供给冲击,其次是价格加成冲击、投资效率冲击和永久性技术冲击。在新的设定下,理论方差分解结果表明,尽管货币政策仍然是非常重要的因素,但是其重要程度已经降低;此外,投资的波动主要是源于投资效率冲击而不是投资品价格冲击。这些发现与国内外现有的结论并不一致。这表明,选择何种方式测度货币政策冲击对于分析中国宏观经济波动至关重要。

In order to analyze China＇s macroeconomic fluctuations, the Bayesian method can be used to estimate a dynamic stochastic general equilibrium model containing various impacting and frictional factors. Such models usually use interest rate rules to measure the impact of monetary policy; however, the monetary quantity rule may be more suitable for China＇s reality. Under the new setting of using the monetary quantity rule, the results of the historical variance decomposition show that during 1992-2016, the main driving factors for the fluctuation of output growth rate are the impact of money supply, the investment shocks and the exogenous demand shocks, of which the two major fluctuations need to be explained by other shocks. The main driving factors of fluctuations in inflation are the money supply shock, followed by the price-plus shock, the investment efficiency shock and the permanent technical shock. Under the new setting, the results of the theoretical variance decomposition show that although monetary policy is still a very important factor, the degree of its importance has been reduced; in addition, the investment fluctuation is mainly due to the impact of investment efficiency rather than the impact of investment product prices. These findings are inconsistent with the existing conclusions both at home and abroad. This shows that the choice of ways to measure the impact of monetary policy is crucial to the analysis of China＇s macroeconomic fluctuations.