我国产出缺口估计及其对通货膨胀预测--两种半结构方法的比较

摘要

This paper describes principal procedures to estimate unobserved variables by Kalman filter and Hodrick Prescott multivariate filter. Furthermore, the authors show that the two approaches are closely linked, and specifically, they explain how to reproduce the Hodrick Prescott multivariate filter using the Kalman filter. As an application, the two approaches are used to estimate China’s output gap based on the time series data from the 1st quarter of 1995 to the 2th quarter of 2012, as well as compare and evaluate the results of the estimation. The results suggest that the output gap can fit the economic fluctuations historically very well, and it is a good indicator of inflationary pressures in China. Conse-quently, the output gap provides a useful signal to the monetary authority.%　　分析Kalman滤波和多变量HP滤波的原理，以及两者之间的内在逻辑联系。以附加预期的菲利普斯曲线为理论基础，利用Kalman滤波和多变量HP滤波方法估计我国1995年第1季度至2012年第2季度的潜在产出和产出缺口，估计结果表明各产出缺口基本上与我国经济运行状况相吻合。利用Coe和McDermott(1997)预测模型和VAR模型检验表明产出缺口能够很好地预测通货膨胀率，尤其是Kalman滤波估计的产出缺口预测效果更好。因此产出缺口对于货币当局的政策分析具有十分重要的意义。