Factor Models and Covariance Matrices
摘要
Factor modelsFactor models involve a reduction of parameters in the covariance matrix, and they are highly related to principal componentsPrincipal components. Since these models have played a very important role in data analysis, we not only introduce classical factor modelsFactor models but also include four extended factor modelsFactor models, namely approximate factor modelsApproximate factor (AF) model, factor-augmented regression modelsFactor-augmented regression (FAR) model, dynamic factor modelsDynamic factor model (DFM), and matrix factor modelsMatrix factor models. In addition, three examples are presented to briefly illustrate empirical applications.