EcoSta 2018: Registration
View Submission - EcoSta2018
A0273
Title: Testing equality of principle components in factor models Authors:  Ningning Xia - Shanghai University of Finance and Economics (China) [presenting]
Jianqing Fan - Princeton University (United States)
Yingying Li - Hong Kong University of Science and Technology (Hong Kong)
Xinghua Zheng - HKUST (China)
Abstract: A test is developed for structural breaks in factor models. We focus on the null hypothesis that the size and direction of factor are constant over time. Because structural change can happen to either the factor loadings or the dynamic process of factors or both, we can not only focus on the change of factor loadings. Based on the fact that the presence of a structural change in factor model yields a size or direction change of principle components obtained from the sample covariance matrix, we test the equality of sample principle components and compare the pre and post break subsample covariance matrix. Our test is consistent under the alternative hypothesis in which a fraction of or all factors have structural changes. The simulation studies show that our test has good finite-sample size and power.