关键词: 数据修订, 最终数据, 实时数据, 泰勒规则

Abstract: This paper extends to build three models, such as backward-looking, synchronism and forward-looking time-varying Taylor rule, then uses them to estimate the time-varying monetary policy of China based on ex-post data and real-time data, and final investigates the influences of data-revision and real-time estimation on the Taylor rule. The estimated results show that the effects of data-revision on the Taylor rule estimation depends on the choice of the models, furthermore, the time-varying parameters of targeting to inflation and output gap are all influenced by data-revision in the three models. In particular, it’s suitable for the ex-post data to choose synchronism Taylor rule model, while the backward-looking Taylor rule model is much better for the real-time data. Finally, this paper puts forward the corresponding suggestions of model selection based on different data.

Key words: Data Revision, Ex-post Data, Real-time Data, Taylor Rule