B1927
Title: Inference on three-pass regression filter with high-dimensional target variables
Authors: Shou-Yung Yin - National Taipei University (Taiwan) [presenting]
Abstract: A framework is considered for high-dimensional target variables using the three-pass regression filter~(3PRF). We propose an estimator that involves two steps. First, we use the diversified projection to extract the information from the high-dimensional target variables. Then we adopt 3PRF to ensure that the factors from regressors can improve the forecast performance. The advantage of this approach is that we do not need to impose a number of factors, and the closed-form solution is easy to obtain. Consistency and asymptotic normality are then established. The simulation study shows that the proposed approach performs well when the number of factors is wrongly imposed while the results of using principal components analysis are sensitive. In the empirical study, we use the proposed method to extract the common components which can be used to predict the fundamentals of the dynamics of house prices in the U.S.