中国通胀预期测度及其对实际通胀的影响——基于贝叶斯Gibbs抽样算法的状态空间模型估计

China's Inflation Expectation Measurement and Its Influence on Real Inflation——Bayesian Estimation for State Space Model Based on Gibbs Sampling Algorithm

通货膨胀预期是影响实际通胀的重要变量,也是货币政策有效运用的关键因素。在新凯恩斯混合菲利普斯曲线的理论框架下构建状态空间模型,利用贝叶斯Gibbs抽样算法估计我国2001—2013年的季度预期通胀率,进一步利用VAR模型及脉冲响应函数分析我国通胀预期对实际通胀的动态影响,实证结果显示:我国季度预期通胀率的适应性特征强于理性特征;适应性预期冲击在短期对实际通胀会产生较大影响,但累积效应在大约9个季度之后消失;理性预期冲击对实际通胀的正向影响会持续较长时间,并最终将实际通胀推高到一个新的水平。因此,货币政策应从降低适应性通胀惯性和管理理性通胀预期两个方面来调控通胀预期对实际通胀的影响。

Inflation expectation is an important variable to affect real inflation and is also the key factor to effectiveapplication of monetary policy. State Space Model is constructed under Keynesians hybrid Phillips Curve Theoryframework. The seasonal expectation inflation rate during 2001-2013 of China is estimated by Bayesian GibbsSampling Estimation Method, and the dynamic influence of China’s inflation expectation on real inflation is furtheranalyzed by VAR Model and Impulsive Response Function. Empirical results show that the suitability feature ofChina’s seasonal expectation inflation rate is stronger than rational feature, that the suitability expectation shockcan exert big impact on real inflation in a short term, and that the accumulative effect disappears after about nineseasons. The results also show that the positive influence of rational expectation shock on real inflation can lastrelatively longer time and will finally push real inflation to a new high level. Thus, monetary policy should regulatethe influence of inflation expectation on real inflation from such two aspects as the decrease of suitability inflationinertia and administrative inflation expectation.