Speaker: 王霞 (Xia Wang)

Title: Estimation and Inference on State-Varying FAVAR Models

Schedule: April 26, Wed 12:00-13:30 PM

Location: 诚明楼（Chengming Hall） RM 315 

 

Introduction: 王霞, 中国人民大学经济学院副教授。主要研究兴趣包括：理论计量经济学，时间序列分析，混频模型及其宏观应用。曾在 Journal of Econometrics、International Economic Review、Econometric Theory、 Journal of Business & Economic Statistics、《经济研究》、《金融研究》、《管理科学学报》等国内外核心期刊发表论文20余篇。

 

Abstract: We develop a nonparametric estimation method and inferential theory for the state-varying factor-augmented vector autoregressive (FAVAR) models, in which both the factor loadings and the FAVAR coefficients vary as general functions of a stochastic state process. Our two-stage estimation procedure combines the local principal component analysis and nonparametric kernel regressions, and we establish the limiting distribution theory for the estimators of the state-varying FAVAR coefficients. We also study the construction of impulse response functions (IRFs) in the state-varying FAVAR models and show the consistency of the estimators of IRFs. In addition, we propose three test statistics to gauge the possible sources of state-dependency features in the underlying FAVAR model. The simulation results indicate that our estimators and test statistics perform well in finite samples. In the empirical study, we apply the state-varying FAVAR model to investigate the monetary policy transmission mechanism and IRFs using the U.S. data.