中国实时金融状况指数的另一种测度--基于函数型数据分析方法

Another Measurement of Chinese Real-Time Financial Condition Index--Based on Functional Data Analysis Method

本文使用28个变量的混频数据集,借助函数型数据建模思想,从连续、动态的视角测度了中国实时金融状况指数。研究表明:金融状况指数(Financial ConditionsIndex,FCI)各组成变量的权重具有时变特征,其大小存在较大差异;样本期内新增贷款是影响FCI的最主要因素,金融危机后货币供应、汇率维度的影响程度日益加深;FCI波动具有明显的周期性特征,在经济繁荣时期FCI处于上升态势,在经济萧条时期FCI处于下落态势,其波动转折态势与中国金融市场主要历史事件相吻合;所构建的FCI对于通货膨胀的代理变量CPI具有先导作用,可以有效预测未来6个月内的通货膨胀趋势。较于现有研究,本文构建的中国实时金融状况指数的相对优势在于能够测度任意时点金融运行状况的静态水平,并且其对通货膨胀的预测效果更好。

This paper measures the real-time Financial Conditions Index(FCI)form the continuous and dynamic perspective within the framework of Functional Data Analysis(FDA)based on the mixed-frequency dataset consisting of 28 selected financial variables.The results show that:firstly,the weights of the component variables of the Financial Conditions Index(FCI)have time-varying characteristics and their magnitudes vary widely;new loans during the sample period are the most important factor affecting FCI,and the degree of influence of the money supply and exchange rate dimensions has deepened after the financial crisis;FCI fluctuations have obvious cyclical characteristics,with FCI in an upward trend during economic boom periods and in a downward trend during economic recessions,and its fluctuations and transitions coincide with major historical events in China's financial markets;the constructed FCI has a leading role for the CPI,a proxy variable for inflation,and can effectively predict the inflation trend in the next six months.Compared with existing studies,the relative advantage of the real-time financial conditions index constructed in this paper is that it can measure the static level of financial performance at any point in time,and it has a better predictive effect on inflation.