中国通货膨胀的动态特征研究

发布时间：2019-07-03 20:16

【摘要】：现有测度通胀惯性的常用模型是仅含自身滞后因子的自回归模型,该模型不能有效反映通胀预期和波动性对通胀惯性的影响。本文在自回归模型基础上,构建了一个包含通胀惯性、学习型预期和波动性特征的通胀动态模型。该动态模型从均值和波动项两个方面反映了我国通胀水平的动态变化趋势和特征。针对上述模型不能有效反映通胀状态体制变化的缺点,本文还引入Markov机制转换模型来测度我国通胀水平的状态转移特征。利用分位数回归方法下的自回归模型、本文构建的通胀动态模型以及Markov机制转化模型实证研究了我国通胀的动态特征。分析表明:我国通胀水平具有较强的惯性特征;通胀惯性的形成机理比较复杂,学习型预期只能部分解释通胀惯性;通胀水平的状态转移时间比较长;通胀水平与其波动性有着正向的关系。
[Abstract]:The existing autoregression model for measuring inflation inertia is an autoregression model with only its own lag factor, which can not effectively reflect the influence of inflation expectations and volatility on inflation inertia. On the basis of autoregression model, this paper constructs an inflation dynamic model which includes inflation inertia, learning expectation and volatility. The dynamic model reflects the dynamic trend and characteristics of inflation level in China from two aspects: mean value and fluctuation term. In view of the shortcomings that the above model can not effectively reflect the change of inflation state system, this paper also introduces the Markov mechanism transition model to measure the state transition characteristics of inflation level in China. By using the autoregression model under the quantile regression method, the inflation dynamic model and the Markov mechanism transformation model constructed in this paper are used to empirically study the dynamic characteristics of inflation in China. The analysis shows that the inflation level in China has strong inertia characteristics; the formation mechanism of inflation inertia is more complex, learning expectation can only partly explain inflation inertia; the state transition time of inflation level is relatively long; and there is a positive relationship between inflation level and its volatility.
【作者单位】：南京大学经济学院;安徽财经大学金融学院;
【基金】：国家社会科学基金项目(11CJY080) 江苏省优势学科建设工程的资助
【分类号】：F822.5;F224

【参考文献】