中科院数学与系统科学研究院

科研进展与学术交流报告会

（第6期）

报告人： 洪永淼 研究员（系统科学研究所研究所 ）

题  目：Econometric Analysis of Time-Varying Models via Interdisciplinary Approaches

时  间：2022.09.23(星期五), 10:40-13:00

地  点：数学院南楼N204室 / 腾讯会议401-896-631

摘 要: Characterization and inference for structural changes has been a longstanding problem in econometrics and statistics. In this talk, some recent developments in econometrics of time-varying models will be discussed, including (1) detection for structural changes of large factor models via discrete Fourier transform, (2) robust forecasting under structural changes via time-varying model averaging, and (3) regularized high-dimensional time-varying generalized method of moments (GMM) estimation via ridge fusion penalization. Applications to macroeconomic forecasts and asset pricing will be considered, and some directions for future research will be discussed.