EcoSta 2019: Start Registration
View Submission - EcoSta2019
A0157
Title: On searching for valid instruments in high-dimensional time-series models Authors:  Cy Sin - National Tsing Hua University (Taiwan) [presenting]
Abstract: With the prevalence of the so-called big data, structural models/equations are often estimated with high-dimensional instruments. That said, research papers in the literature either (1) assumes all instruments are valid and considers an efficient estimator; or (2) proposes some confidence sets of the structural parameters, and investigates their properties under various assumptions on the number of valid instruments. We adopt and modify the OGA-HDAIC approach and search for valid instruments out of some high-dimensional potential instruments. Unlike Lasso, this algorithm is arguably more suitable for time-series data. We close with (i) Some comparisons with the high-dimensional Durbin-Wu-Hausman (DWH) test; and (ii) Some Monte-Carlo simulations.