FCI可以作为货币政策的良好指示器吗——基于信息预测检验与工具变量选择的分析

Could FCI Serve as A Good Indicator of Monetary Policy?——An Analysis Based on the Information Prediction Test and Instrumental Variables Selections

本文首先构建了一个包含利率、汇率、房价、股价和货币供应量的五因素FCI,对其进行信息预测性检验,然后将FCI纳入工具变量信息集并对前瞻性泰勒规则进行估计。研究发现,五因素FCI能够较准确地预测未来通胀变化,而对GDP增速的预测效果稍显逊色。Hansen过度识别检验证实,FCI显著提高了工具变量的有效性,且包含FCI的泰勒规则是经验合意的。本文建议货币当局构建一个可有效监测金融市场松紧程度的金融形势指数,同时继续推进利率市场化,为货币政策调控工具向价格型转型奠定基础。

This paper first builds a five-factor financial condition index （FCI） including interest rates, exchange rates, housing prices, stock prices and money supply and then conducts a predictive information test for it. Then we estimate the forward-looking Taylor＇s Rule by taking FCI into the instrumental variable information. This study finds this five factor FCI could accurately predict the future changes in inflation, though its forecast for GDP growth is less effective than that of inflation. Hanson excessive identification test verifies the FCI can significantly improve the effectiveness of instrumental variables and the Taylor＇s Rule including the FCI is experimentally acceptable. This paper suggests the monetary authority should build a FCI which can monitor the tightness of financial markets effectively, as well as continue to promote the reform of market-oriented interest rates to lay the foundation for the transformation of conducting monetary policy through price-based instruments.