A0322
Title: Matrix-valued factor model with time-varying main effects
Authors: Clifford Lam - London School of Economics and Political Science (United Kingdom) [presenting]
Zetai Cen - University of Bristol (United Kingdom)
Abstract: The matrix-valued time-varying main effects factor model (MEFM) is introduced. MEFM is a generalization to the traditional matrix-valued factor model (FM). Rigorous definitions of MEFM and its identifications are given, and estimators are proposed for the time-varying grand mean, row and column main effects, and the row and column factor loading matrices for the common component. Rates of convergence for different estimators are spelled out, with asymptotic normality shown. The core rank estimator for the common component is also proposed, with the consistency of the estimators presented. A test is proposed for testing if FM is sufficient against the alternative that MEFM is necessary, and the power of such a test is demonstrated in various simulation settings. Numerically, the accuracy of the estimators is demonstrated in extended simulation experiments. A set of NYC taxi traffic data is analyzed, and the test suggests that MEFM is indeed necessary for analyzing the data against a traditional FM.
