A0607
Title: Inference and automatic instrument selection in a semiparametric single-index conditional factor model
Authors: Qi Zhang - Southwestern University of Finance and Economics (China) [presenting]
Qiuhua Xu - Southwestern University of Finance and Economics (China)
Chen Huang - Aarhus University (Denmark)
Abstract: A semi-parametric conditional latent factor model with a single-index structure is investigated for estimation and variable selection. The sieve method and singular value decomposition are employed to estimate latent factors, while adaptive LASSO is utilized to identify relevant characteristics. Both large-N and large-NT asymptotics are established for all estimators and the pricing-error test. Variable selection exhibits consistency and Oracle properties. A weighted bootstrap procedure is developed to test the null hypothesis that pricing errors are zero. Empirical analysis of U.S. equity data illustrates out-of-sample predictive performance and the set of selected characteristics.
