我国通货膨胀的高频和低频解释——基于DWT的频段回归分析

Interpreting Chinese Inflation at High and Low Frequency——DWT-based Spectral Regression Analysis

利用我国1993-2010年季度数据,以频段回归方法分析货币增长和实际产出缺口等因素在不同频段与通货膨胀的关系。结果显示,在高频(短期),实际产出缺口与通胀显著线性相关;在低频(长期),货币量与通胀显著线性相关。结论表明,实体经济短期波动对我国短期通胀影响显著,货币量变化对我国长期通胀影响显著。因此,实体经济短期波动分析有利于通胀短期预测,货币量分析有利于通胀长期预测,二者结合可以提高央行通胀预测的准确性,从而有利于物价稳定目标的实现。

With 1993- 2010 quarterly data,the paper analyses the relationship between inflation and money aggregate,real output gap with spectral regression. The result shows that at high frequency real output gap is significantly correlated with inflation,while at low frequency money aggregate is significantly correlated with inflation. This indicates that real economy fluctuation analysis is useful for central bank to anticipate inflation in the short term,while money aggregate analysis is useful for central bank to anticipate inflation in the long term. Therefore,real economy fluctuation analysis and money aggregate analysis are all important for inflation anticipation. It is helpful for central bank to apply both analyses in managing inflation accurately.