我国新型金融状况指数的构建与物价预测

发布时间：2018-06-22 20:44

  本文选题：金融状况指数(FCI) + 通胀率　； 参考：《财经问题研究》2017年06期

【摘要】：本文利用时变因子载荷矩阵TVP-FAVAR模型提取出相关金融变量的共同因子,并借助动态模型选择方法筛选每期的最优因子,以此构建了我国新型金融状况指数(FCI),从相关性、因果关系和预测能力等视角分析了FCI与通胀率之间的关系。结果显示:以时变参数和时变维度方法构建的新型FCI与我国宏观经济现实吻合程度较好;相对于传统方法,本文构建的FCI与通胀率之间具有较高的相关性和较低的预测误差,且FCI相对通胀率平均先行2-3个月,能够较好地预测通胀的短期未来走势,可以作为货币政策制定的参考指标。
[Abstract]:In this paper, we use the time-varying factor of factor load matrix TVP-FAVAR model to extract the common factors of relevant financial variables, and select the optimal factor of each period with the dynamic model selection method. In order to build a new financial situation index (FCI) in China, the relationship between FCI and the rate of inflation is analyzed from the angle of relevance, causality and predictive energy. The results show that the new FCI constructed with time-varying parameter and time-varying dimension method is in good agreement with our macro-economic reality. Compared with the traditional method, the FCI and the inflation rate have high correlation and low prediction error, and the relative inflation rate of FCI is 2-3 months on average, which can better predict the short term inflation. The trend can be used as a reference index for monetary policy making.
【作者单位】： 东北财经大学经济学院;东北财经大学经济计量分析与预测研究中心;
【基金】：国家社会科学基金重大项目“新常态下我国宏观经济监测和预测研究”(15ZDA011) 国家自然科学基金项目“基于非参数方法和非线性模型的经济景气和通货膨胀监测预警研究”(71173029) 辽宁省特聘教授(2012)项目
【分类号】：F822.5;F832
，

本文编号：2054174