基于VAR的我国预期通货膨胀率的估计问题研究

Study on the Estimation of Inflation Expectation of China Based on VAR Model

在当前影响物价变动的因素愈加广泛和复杂的背景下,如何准确把握未来通货膨胀预期走势进而有效调控通货膨胀至关重要。本文首先通过建立附加前瞻性政策变量的VAR预期模型,根据2002年第一季度至2014年第三季度的通货膨胀率、实际利率和产出偏差的实际值与预期目标值的偏差值季度数据,采用卡尔曼滤波递归算法得出我国的通货膨胀预期的估计结果。随后基于理性预期理论对初步估计结果进行检验。研究结果表明,采用前三个月实际利率、实际通货膨胀率的算术平均作为下期中国人民银行调控目标的预期值是较为符合我国情况的选择。

Under the background that the factors affecting the price changes are more extensive and complicated, it is very important to accurately grasp the future inflation expectation and effectively control the inflation. According to the deviation values of the inflation rate, real interest rate and output deviation from 2002 Q1 to 2014 Q3, this paper firstly establishes the VAR expectation model with additional predictive policy variables, and uses the Kalman filter recursive algorithm to obtain the estimate result of Chinese inflation expectations. Then the preliminary estimation results are tested based on the rational expectation theory. The results show that using the arithmetic average of the real interest rate and actual inflation rate in the first three months as the expected value is very suitable to set the regulation target of the People＇s Bank of China in the next period.